Welcome to ISMB 2020, the flagship meeting of ISCB!

Welcome to the ISMB 2020 virtual platform. The International Society for Computational Biology (ISCB) is grateful for your support and participation as we navigate these unprecedented times. We, like you, were devastated when we had to cancel our face to face conference, but we knew that we needed to find a way to continue to advance computational research and provide our members with a platform to disseminate research findings and network.

Over the course of the next four days, you will have the opportunity to hear and interact live with speakers presenting over 400 talks in 22 Community of Special Interest (COSI) tracks, Special Sessions, Technology Tracks, and a Workshop on Bioinformatics Education (WEB). You will also be able to browse more than 700 posters in the virtual Poster Hall, where authors will be standing by to answer your questions in the chat feature. And don’t forget to stop by our sponsorship and exhibitor section to learn more about publishing opportunities, services, tools, and job openings.

Schedule

All times are Pacific

Sort by Date

University of Delaware

10:40 AM - 11:20 AM (EDT)

iRNA - iRNA Keynote: The Functional Iso-Transcriptomics toolset to leverage long reads sequencing for unraveling isoform transcriptional networks from single cells.

Long read sequencing is revolutionizing our ability to study transcriptome diversity and dynamics by empowering the detection of full-length transcripts. However, long reads pose novel bioinformatics challenges both for removing technology biases as for leveraging the discovery potential of the new data. In this talk I will present the Functional Iso-Transcriptomics analysis toolset composed of SQANTI, IsoAnnot and tappAS, that covers quality control, annotation and statistical analysis of long reads transcriptome data. I will also illustrate how to use these tools to unravel isoform transcriptional networks combining long and short single cell RNA-seq data.

Speakers:

Ana Conesa

University of Florida

10:40 AM - 11:40 AM (EDT)

Function - Function COSI Keynote: DNA methylases – computation, experiments and new biology

Speakers:

Richard Roberts

10:40 AM - 11:40 AM (EDT)

MLCSB: Machine Learning - Emergent pathogens, vaccines and therapeutics : how can computation accelerate discovery?

Speakers:

Oliver Stegle

10:45 AM - 11:40 AM (EDT)

CAMDA - CAMDA KEYNOTE: Climate, Oceans, and Human Health: Cholera as a paradigm for prediction infectious diseases

Climate and the oceans historically have been closely intertwined with human health. Today significant advances in information technology have brought new discoveries - from the outer reaches of space, where remote sensing monitors on satellites circle the earth, to the ultramicroscopic through application of next generation sequencing and bioinformatics. Vibrio cholerae provides a useful example of the fundamental link between human health and the oceans. This bacterium is the causative agent of cholera and is associated with major pandemics, yet it is a marine bacterium with a versatile genetics and is distributed globally in estuaries throughout the world, notably the Bay of Bengal, but also in coastal regions and aquatic systems of the world. Vibrio species, both nonpathogenic and those pathogenic for humans, marine animals, or marine vegetation, play a fundamental role in nutrient cycling. They have also been shown to respond to warming of surface waters of the North Atlantic, with increase in their numbers correlated with increased incident of vibrio disease in humans. The models we have developed for understanding and predicting outbreaks of cholera are based on work done in the Chesapeake Bay and the Bay of Bengal and these models are now used by UNICEF and aid agencies to predict cholera in Yemen and other countries of the African continent. With onset of COVID-19, these models are currently being modified to predict SARS CoV-2 and incidence of COVID-19, the current pandemic of coronavirus. In summary, molecular microbial ecology coupled with computational science can provide a critical indicator and prediction of human health and wellness. How this is being accomplished and how we are beginning to understand environmental aspects of COVID-19 will be discussed in this talk.

Speakers:

University of Florida

11:10 AM - 11:40 AM (EDT)

SST01 - The impact of cancer associated CTCF mutations on chromatin architecture and gene regulation

Speakers:

Jane Skok

New York University School of Medicine

11:20 AM - 11:40 AM (EDT)

Breakout session 1

We will break into small groups for discussion of topics TBA.

Speakers:

Center for Algorithmic Biotechnology, Institute of Translational Biomedicine, St. Petersburg State University

12:00 PM - 12:20 PM (EDT)

Breakout session 2

We will break into small groups for discussion of topics TBA.

Speakers:

Multiple Authors

12:00 PM - 12:10 PM (EDT)

Text Mining - BERTMeSH: Deep Contextual Representation Learning for Large-scale High-performance MeSH Indexing with Full Text

large-scale automatic Medical Subject Headings (MeSH) indexing has become increasingly important. FullMeSH, the only method for large-scale MeSH indexing with full text, suffers from three major drawbacks: FullMeSH 1) uses Learning To Rank (LTR), which is time-consuming, 2) can capture some pre-defined sections only in full text, and 3) ignores the whole MEDLINE. We propose a deep learning based method, BERTMeSH, which is flexible for section organization in full text. BERTMeSH has two technologies: 1) the state-of-the-art pre-trained deep contextual representation, BERT (Bidirectional Encoder Representations from Transformers), which makes BERTMeSH capture deep semantics of full text. 2) a transfer learning strategy for using both full text in PubMed Central (PMC) and title and abstract in MEDLINE, to take advantages of both. In our experiments, BERTMeSH was pre-trained with 3 million MEDLINE citations and trained on approximately 1.5 million full text in PMC. BERTMeSH outperformed various cutting edge baselines. For example, for 20K test articles of PMC, BERTMeSH achieved a Micro F-measure of 69.2%, which was 6.3% higher than FullMeSH with the difference being statistically significant.Also prediction of 20K test articles needed 5 minutes by BERTMeSH, while it took more than 10 hours by FullMeSH, proving the computational efficiency of BERTMeSH.

Speakers:

Ronghui You

Fudan University

12:00 PM - 12:10 PM (EDT)

iRNA - Full-length transcriptome reconstruction reveals a large diversity of RNA and protein isoforms in rat hippocampus

Gene annotation is a critical resource in genomics research. Many computational approaches have been developed to assemble transcriptomes based on high-throughput short-read sequencing, however, only with limited accuracy. Here, we combine next-generation and third-generation sequencing to reconstruct a full-length transcriptome in the rat hippocampus, which is further validated using independent 5 ́ and 3 ́-end profiling approaches. In total, we detect 28,268 full-length transcripts (FLTs), covering 6,380 RefSeq genes and 849 unannotated loci. Based on these FLTs, we discover co-occurring alternative RNA processing events. Integrating with polysome profiling and ribosome footprinting data, we predict isoform-specific translational status and reconstruct an open reading frame (ORF)-eome. Notably, a high proportion of the predicted ORFs are validated by mass spectrometry-based proteomics. Moreover, we identify isoforms with subcellular localization pattern in neurons. Collectively, our data advance our knowledge of RNA and protein isoform diversity in the rat brain and provide a rich resource for functional studies.

Speakers:

Xi Wang

German Cancer Research Center

12:00 PM - 12:20 PM (EDT)

Function - Update on Protein Functional Annotation in UniProt in 2020

With the increasing number of data generated by sequencing projects, researchers need reliable systems to provide the functional annotation of proteins. UniProtKB is using two functional annotation systems, UniRule and ARBA (Association-Rule-Based Annotator), to automatically annotate UniProtKB/TrEMBL in an efficient and scalable manner with a high degree of accuracy. These systems use protein signatures and taxonomy classifications to infer the biochemical features and biological functions of proteins. This knowledge is expressed in the form of rules: sets of IF-THEN statements coming from expert curation (UniRule [1]) or generated by machine learning (ARBA [2]). Rules are applied at each release keeping the propagated annotations up-to-date. On the UniProtKB website, information added by annotation rules is clearly highlighted as such using evidence tags. These tags can also be used as keywords to search for or filter out annotation added by a rule. The protein function community could also benefit from those rules, as some sequences may not be yet available in public databases or could be present in highly redundant proteomes absent from UniProtKB. This has been made possible via UniFIRE (The UniProt Functional annotation Inference Rule Engine), an engine to execute rules in the URML (UniProt Rule Markup Language) format.

Speakers:

Rabie Saidi

European Molecular Biology Laboratory, European Bioinformatics Institute (EMBL-EBI)

12:00 PM - 12:20 PM (EDT)

MLCSB: Machine Learning - Cancer mutational signatures representation by large-scale context embedding

The accumulation of somatic mutations plays critical roles in cancer development and progression. However, the global patterns of somatic mutations, especially non-coding mutations, and their roles in defining molecular subtypes of cancer have not been well characterized due to the computational challenges in analyzing the complex mutational patterns. Here we develop a new algorithm, called MutSpace, to effectively extract patient-specific mutational features using an embedding framework for larger sequence context. Our method is motivated by the observation that the mutational rate at megabase scale and the local mutational patterns jointly contribute to distinguishing cancer subtypes, both of which can be simultaneously captured by MutSpace. Simulation evaluations demonstrate that MutSpace can effectively characterize mutational features from bona fide patient subgroups and achieve superior performance compared with previous methods. As a proof-of-principle, we apply MutSpace to 560 breast cancer patient samples and demonstrate that our method achieves high accuracy in subtype identification. In addition, the learned embeddings from MutSpace reflect intrinsic patterns of breast cancer subtypes and other features of genome structure and function. MutSpace is a promising new framework to better understand cancer heterogeneity based on somatic mutations.

Speakers:

Yang Zhang

Carnegie Mellon University

12:00 PM - 12:40 PM (EDT)

CAMDA - Metagenomic Geolocation using Read Signatures

We present a novel approach to the Metagenomic Geolocation Challenge based on random projection of the sample reads for each location. Individual signatures are computed for all reads, and each location is characterised by an hierarchical vector space representation of the resulting clusters. Classification is then treated as a problem in ranked retrieval of locations, where similar signatures are taken as a proxy for underlying microbial similarity. We evaluate our approach based on the 2016 and 2020 Challenge datasets and obtain promising results based on nearest neighbour classification.

Speakers:

Dimitri Perrin

Queensland University of Technology

12:00 PM - 12:40 PM (EDT)

SysMod - Robotic mapping and generative modelling of cytokine response

We have developed a robotic platform allowing us to monitor cytokines dynamics (including IL-2, IFNgamma, TNF alpha, IL-6) of immune cells in vitro, with unprecedented resolution. To understand the complex emerging dynamics, we use interpretable machine learning techniques to build a generative model of cytokine response. We discover that, surprisingly, immune activity is encoded into one global parameter, encoding ligand antigenic properties and to a less extent ligand quantity. Based on this we build a simple interpretable model which can fully explain the broad variability of cytokines dynamics. We validate our approach using different lines of cells and different ligands. Two processes are identified, connected to timing and intensity of cytokine response, which we successfully modulate using drugs or by changing conditions such as initial T cell numbers. Our work reveals a simple "cytokine code", which can be used to better understand immune response in different contexts including immunotherapy. More generally, it reveals how robotic platforms and machine learning can be leveraged to build and validate systems biology models. Work in collaboration with Grégoire Altan-Bonnet, National Cancer Institute

Speakers:

Paul Francois

McGill University

12:00 PM - 12:40 PM (EDT)

SST01 - Probing the Immune Response to Vaccination with Systems Biology

Speakers:

Bali Pulendran

Stanford University

12:10 PM - 12:20 PM (EDT)

iRNA - Freddie: Annotation-free Isoform Discovery Using Long-Read Sequencing

Background Alternative splicing (AS) events are essential to understanding the development of cancer and may play a role as a target of personalized cancer therapeutics. However, detecting novel AS events remains a challenging task: existing reference transcriptome annotation databases are far from universal comprehensiveness and traditional sequencing technologies are limited by their short-read lengths. Given these challenges, transcriptomic long-read sequencing (LRS) presents a promising potential for novel AS discovery. Methods We present Freddie, a computational annotation-free isoform discovery and detection tool. Freddie uses genome alignments of transcriptomic LRS reads as input and generates isoform clusters of these reads for a given gene of interest. Freddie then segments the gene interval using the alignments into canonical exon segments. Finally, Freddie clusters the reads into isoform clusters that satisfy a set of expected transcriptomic LRS constraints. We formulate this clustering as an optimization problem that we name Minimum Error Clustering into Isoforms (MErCi) problem and solve it an Integer Linear Program. Results We compare the performance of Freddie on both simulated and real datasets with state-of-the-art isoform detection tools with varying dependence on reference annotations. We show that both the segmentation and clustering steps of Freddie are highly accurate and computationally efficient.

Speakers:

Baraa Orabi

The University of British Columbia

12:10 PM - 12:20 PM (EDT)

Text Mining - Integrating Image Caption Information into Biomedical Document Classification in Support of Biocuration

Biological databases provide precise, well-organized information for supporting biomedical research. Developing such databases typically relies on manual curation. However, the large and rapidly increasing publication rate makes it impractical for curators to quickly identify all and only those documents of interest. As such, automated biomedical document classification has attracted much attention. Images convey significant information for supporting biomedical document classification. Accordingly, using image caption, which provides a simple and accessible summary of the actual image content, has potential to enhance the performance of classifiers. We present a classification scheme incorporating features gathered from captions in addition to titles-and-abstracts. We trained/tested our classifier over a large imbalanced dataset, originally curated by Gene Expression Database (GXD), consisting of ~60,000 documents, where each document is represented by the text of its title, abstract and captions. Our classifier attains precision of 0.698, recall 0.784, f-measure 0.738 and Matthews correlation coefficient 0.711, exceeding the performance of current classifiers employed by biological databases and showing that our framework effectively addresses the imbalanced classification task faced by GXD. Moreover, our classifier’s performance is significantly improved by utilizing information from captions compared to using titles-and-abstracts alone, which demonstrates that captions provide substantial information for supporting biomedical document classification.

Speakers:

Xiangying Jiang

University of Delaware

12:20 PM - 12:30 PM (EDT)

iRNA - Prioritizing genes likely to have functionally distinct splice isoforms using long read RNA-seq data

RNA sequencing continues to detect an increasing number of splice variants, yet there are limited genes with evidence of functionally distinct splice isoforms (FDSIs). One challenge is that splice variants are often reconstructed computationally from short sequence fragments, an error-prone process. Here we identify likely candidate genes with FDSIs using long-read RNA-sequencing data that removes much of the present ambiguity around transcript structures. We developed a computational pipeline for prioritizing splice isoforms in MinION long-read RNA-seq data. We designed the prioritization approach using splice variant-specific conservation (PhastCons, PhyloP and BLAST homology searches), expression, coding-potential, and protein domain annotations. Based on these annotations, the pipeline outputs a prioritized list of genes likely to have FDSIs. We then applied this pipeline to publicly available and novel mouse brain and liver transcriptomes. For 6,799 genes with multiple splice variants, our approach prioritized a set of 44 putative genes with FDSIs. This candidate set includes genes with published evidence of FDSIs (Cdc42) and genes with promising literature evidence of FDSIs (Tpd52l1, Gstz1). The limited amount of long-read data and low sequencing depth hinders our prioritization. Nevertheless, our work aids in establishing guidelines for high-throughput prioritization of genes with FDSIs.

Speakers:

Shamsuddin Bhuiyan

The University of British Columbia

12:20 PM - 12:40 PM (EDT)

Function - Pruning the Protein Jungle: recent developments in the CATH-Gardener function analysis and prediction pipeline.

The CATH-Gene3D database includes evolutionary relationships between protein domains, and classifies domains into superfamilies. Functional Families (FunFams) subdivide superfamilies further to provide clusters and alignments of protein domains that all perform closely related functions. FunFams have performed well in function prediction assessments (CAFA) and provide insights both on functional sites and effects of variants on structure and disease. FunFams are created with Gardener: a novel pipeline for clustering massive protein datasets. Domains are first partitioned by Multi-Domain Architecture (MDA) then tree-building/tree-cutting algorithms (GeMMA/FunFHMMER) are applied iteratively. To deal with huge numbers of sequences from metagenome projects, we have employed a novel random sampling approach that reduces search space, whilst retaining functional diversity required to build informative alignments. For CAFA4, we implemented a generalised Fast and Frugal Tree algorithm to filter and combine search results from a variety of sources. Sources contributing to these results included: FunFams built from CATH and Pfam domains, a predictor based on autoencoding of network information and a Machine Learning (ML) approach trained to recognise patterns in FunFam sequence matches. Benchmarks suggest significant improvement in accuracy for the most recent FunFams, also that network data considerably enhances function prediction in in-silico validation on fission yeast.

Speakers:

Nicola Bordin

University College London

12:20 PM - 12:30 PM (EDT)

Text Mining - Bridging the gap: enlisting authors’ help addressing the remaining difficulties in automated concept extraction

Automated text analysis has proven very effective for retrieving relevant articles from the biomedical literature. But quantitative biomedical research requires the specific knowledge embedded within individual articles and more comprehensive results across the literature. Text mining helps provide this data by automating the conversion of unstructured text such as scientific publications into structured, computable formats. Recent advances allow automated text mining systems to provide results that are predominantly of very high quality. However, some cases remain difficult for current text mining tools. In this work we propose ten writing tips to enlist the help of authors. We also propose a web-based tool, PubReCheck, to help authors visualize results of automated concept extraction on their text and to automatically identify many types of issues prior to publication. Articles that follow these guides will typically be processed more accurately, enabling the content to be found more readily and used more widely. Following these guidelines at a large scale will improve the ability of millions to find articles that meet their information needs and the ability of text mining tools to provide structured, computable data that enable larger scale and more rapid analyses. Availability: https://www.ncbi.nlm.nih.gov/CBBresearch/Lu/PubReCheck/

Speakers:

Robert Leaman

NCBI/NLM/NIH

12:20 PM - 12:40 PM (EDT)

Breakout session 3

Groups report their findings back to the main group.

Speakers:

Multiple Authors

12:20 PM - 12:40 PM (EDT)

MLCSB: Machine Learning - MHCAttnNet: Predicting MHC-Peptide Bindings for MHC Alleles Classes I & II Using An Attention-Based Deep Neural Model

Motivation: Accurate prediction of binding between an MHC allele and a peptide plays a major role in the synthesis of personalized cancer vaccines. The immune system struggles to distinguish between a cancerous and a healthy cell. In a patient suffering from cancer who has a particular MHC allele, only those peptides that bind with the MHC allele with high affinity help the immune system recognize the cancerous cells. Results: MHCAttnNet is a deep neural model that uses an attention mechanism to capture the relevant subsequences of the amino acid sequences of peptides and MHC alleles. It then uses this to accurately predict the MHC-peptide binding. MHCAttnNet achieves an AUC-PRC score of 94.18% with 161 class I MHC alleles which outperforms the state-of-the-art models for this task. MHCAttnNet also achieves a better AUC-ROC score in comparison to the state-of-the-art models while covering a greater number of class II MHC alleles. The attention mechanism used by MHCAttnNet provides a heatmap over the amino acids thus indicating the important subsequences present in the amino acid sequence. This approach also allows us to focus on a much smaller number of relevant trigrams corresponding to the amino acid sequence of an MHC allele, from 9251 possible trigrams to about 258. This significantly reduces the number of amino acid subsequences that need to be clinically tested.

Speakers:

Aayush Grover

International Institute of Information Technology Bangalore

12:30 PM - 12:40 PM (EDT)

Text Mining - Promoting Reproducible Research for Characterizing Nonmedical use of Medications through Data Annotation: A Description of a Twitter Corpus and Guidelines

Due to the contribution of prescription medications, such as stimulants or opioids, to the broader drug abuse crisis, several studies have explored the use of social media data, such as from Twitter, for monitoring prescription medication abuse. However, there is a lack of clear descriptions of how abuse information is presented in public social media, as well as few thorough annotation guidelines that may serve as the groundwork for long-term future research or annotated datasets usable for automatic characterization of social media chatter associated with abuse-prone medications. We employed an iterative annotation strategy to create an annotated corpus of 16,433 tweets, mentioning one of 20 abuse-prone medications, labeled as potential abuse or misuse, non-abuse consumption, drug mention only, and unrelated. We experimented with several machine learning algorithms to illustrate the utility of the corpus and generate baseline performance metrics for automatic classification on these data. Among the machine learning classifiers, support vector machines obtained the highest automatic classification accuracy of 73.00% (95% CI 71.4-74.5) over the test set (n=3271). We expect that our annotation strategy, guidelines, and dataset will provide a significant boost to community-driven data-centric approaches for the task of monitoring prescription medication misuse or abuse monitoring from Twitter.

Speakers:

Karen Oconnor

University of Pennsylvania

12:30 PM - 01:30 PM (EDT)

iRNA - iRNA Panel Discussion: Long-read RNA-seq

Long read discussion going into lunch break

Speakers:

Yoseph Barash

University of Pennsylvania

01:00 PM - 01:45 PM (EDT)

ISCB Town Hall

Speakers:

Diane Kovats

ISCB

Thomas Lengauer

Max Planck Institute for Informatics

02:00 PM - 02:10 PM (EDT)

SST02 - Introduction

This special session aims to introduce the audience to the motivations, principles and opportunities offered by crowdsourcing technologies. We will emphasize the methodological contributions and describe upcoming technical challenges. This talk will present the speakers and organization of the session.

Speakers:

Jérôme Waldispühl

McGill University

02:00 PM - 02:20 PM (EDT)

iRNA - Inferring snoRNA characteristics from their abundance profile in healthy human tissues

Small nucleolar RNAs (snoRNAs) are noncoding RNAs known to regulate ribosome biogenesis and splicing. SnoRNAs have a highly stable structure which impairs their quantification in high-throughput sequencing (RNA-Seq). The use of thermostable group II intron reverse transcriptase in RNA-seq (TGIRT-Seq) was recently shown to accurately quantify their abundance. To faithfully characterize snoRNAs’ abundance determinants, we carried out the TGIRT-Seq of seven healthy human tissues and conducted subsequent bioinformatic analyses. We found that snoRNAs can be categorized in two abundance profiles with distinct characteristics: uniformly expressed (UE) and tissue-specific (TS) snoRNAs. UE snoRNAs are encoded in protein-coding host gene (HG), are highly conserved and target rRNA whereas TS snoRNAs are encoded in noncoding HG, are poorly conserved and are orphan snoRNAs. We found that UE snoRNAs can be anticorrelated with their HG (in which case the HG is mostly involved in ribosome biogenesis and translation) or correlated (in which case the HG is mostly coding for a ribosomal protein) in regards of their abundance. Conversely, TS snoRNAs are well correlated with their HG. These results suggest a model in which, based upon its abundance profile across tissues, a snoRNA’s target, conservation and functional relationship with its HG can be clearly inferred.

Speakers:

Étienne Fafard-Couture

Université de Sherbrooke

02:00 PM - 02:20 PM (EDT)

Function - Benchmarking Gene Ontology Function Predictions Using Negative Annotations

With the ever-increasing number and diversity of sequenced species, the challenge to characterise genes with functional information is even more important. In most species, this characterisation almost entirely relies on automated electronic methods. As such, it is critical to benchmark the various methods. The CAFA series of community experiments provide the most comprehensive benchmark, with a time-delayed analysis leveraging newly curated experimentally supported annotations. However, the definition of a false positive in CAFA has not fully accounted for the Open World Assumption (OWA), leading to a systematic underestimation of precision. The main reason for this limitation is the relative paucity of negative experimental annotations. This paper introduces a new, OWA-compliant, benchmark based on a balanced test set of positive and negative annotations. The negative annotations are derived from expert-curated annotations of protein families on phylogenetic trees. This approach results in a large increase in the average information content (IC) of negative annotations. The benchmark has been tested using the naïve and BLAST baseline methods, as well as two orthology-based methods. This new benchmark could complement existing ones in future CAFA experiments.

Speakers:

Alex Vesztrocy

University of Lausanne

02:00 PM - 02:20 PM (EDT)

CAMDA - Unraveling city-specific microbial signature and identifying sample origin for the data from CAMDA 2020 Metagenomic Geolocation Challenge

As the composition of microbial communities can be location-specific, investigating the microbial community of different cities could unravel city-specific microbial signatures and further identify the origin of future samples. In this study, the whole genome shotgun (WGS) metagenomics data from over 20 cities in 17 countries and city-specific information including location, climate and biomes data were provided as part of the CAMDA 2020 “Metagenomic Geolocation Challenge”. Appropriate methods including feature selection, normalization, two popular machine learning methods (Random Forest (RF) and Support Vector Machine (SVM)), one deep learning method (Multilayer Perception (MLP)), and Principal Coordinates Analysis (PCoA) were used.

Speakers:

Dorothy Ellis

University of Florida

02:00 PM - 02:30 PM (EDT)

SST01 - Chemistry between genes: a proposition for ImmunoMetabolomics

Speakers:

Shuzhao Li

The Jackson Laboratory for Genomic Medicine

02:00 PM - 02:40 PM (EDT)

SysMod - Cross-Species Translation of Biological Information via Computational Systems Modeling Frameworks

A vital challenge that the vast majority of biological research must address is how to translate observations from one physiological context to another—most commonly from experimental animals (e.g., rodents, primates) or technological constructs (e.g., organ-on-chip platforms) to human subjects. This is typically required for understanding human biology because of the strong constraints on measurements and perturbations in human in vivo contexts. Direct translation of observations from experimental animals to human subjects is generally unsatisfactory because of significant differences among organisms at all levels of molecular properties from genome to transcriptome to proteome and so forth. Accordingly, addressing inter-species translation requires an integrated experimental/computational approach for mapping comparable but not identical molecule-to-phenotype relationships. This presentation will describe methods we have developed for a variety of cross-species translation examples, demonstrated on applications in inflammatory pathologies and cancer.

Speakers:

Douglas Lauffenburger

02:00 PM - 02:40 PM (EDT)

Text Mining - Text Mining Keynote: Text mining to understand drug action: from PubMed to Reddit

Speakers:

Russ Altman

Stanford University

02:00 PM - 03:00 PM (EDT)

MLCSB: Machine Learning - Deep learning at base-resolution reveals motif syntax of the cis-regulatory code

Genes are regulated by cis-regulatory elements, which contain transcription factor (TF) binding motifs in specific arrangements. To understand the syntax of these motif arrangements and its influence on TF binding, we developed a new convolutional neural network called BPNet that models the relationship between regulatory DNA sequence and base-resolution binding profiles from ChIP-exo/nexus experiments targeting four pluripotency TFs Oct4, Sox2, Nanog, and Klf4 in mouse embryonic stem cells. BPNet is able to predict base-resolution binding profiles and footprints at unprecedented accuracy on par with replicate experiments. We developed a suite of model interpretation methods to learn novel motif representations, accurately map predictive motif instances in the genome and identify higher-order rules by which combinatorial motif syntax influences binding of these TFs. We discovered several novel motifs bound by these TFs supported by distinct footprints. We further found that instances of strict motif spacing are largely due to retrotransposons, but that soft motif syntax influences TF binding at protein or nucleosome range in a directional manner. Most strikingly, Nanog binding is driven by motifs with a strong preference for ~10.5 bp spacings corresponding to helical periodicity. We then validated our model's predictions using CRISPR-induced point mutations of motif instances followed by profiling TF binding. The sequence representations learned by the binding models also generalized to accurately predict differential chromatin accessibility after TF depletion as well as massively parallel reporter experiments. BPNet easily adapts to other types of profiling experiments (e.g. ChIP-seq, DNase-seq, ATAC-seq, PRO-seq) in mammals as well as other species such as yeast, thus paving the way to decipher the cis-regulatory code from diverse regulatory profiling experiments.

Speakers:

Anshul Kundaje

Stanford University

02:00 PM - 02:04 PM (EDT)

Web - Introduction to WEB 2020

Speakers:

Michelle Brazas

Ontario Institute for Cancer Research

02:04 PM - 02:05 PM (EDT)

Web - RECENT ONLINE TRAINING EXPERIENCES - THE UNIVERSITY EXPERIENCE

Speakers:

Michelle Brazas

Ontario Institute for Cancer Research

02:05 PM - 02:25 PM (EDT)

Web - Teaching the Virus: Lessons from the Online Age

We are five faculty members who teach in Frontiers of Science, the Columbia College science core course. Like most of our colleagues around the world, we turned to online teaching in mid-March. To stay true to our name – Frontiers of Science, we modified our curriculum to make room for teaching about the evolution of SARS-CoV-2 using genomic data. After an introduction on the evolution of coronaviruses and genomics studies, we conducted an online activity, where students worked in teams to investigate the evolution of the SARS-CoV-2 using the Nextstrain.org website. First, students thoroughly examined phylogenetic graphs and discussed how these trees are created. Next, students calculated the virus’ substitution rates per site and compared it to that of the other pathogens, such as Influenza. The class ended with discussions revolving around infectious diseases and the role of genomics in treatment and vaccine development. We hope that other faculty may be able to use or adapt the activity that we have developed on SARS-CoV-2 and some of the ideas that have guided our modified online curriculum. If there ever was a topic that all students would deem to be relevant to their lives, this is certainly one.

Speakers:

Tugce Bilgin

Columbia University

02:10 PM - 02:30 PM (EDT)

SST02 - Foldit and the origin of scientific discovery games in molecular biology

This talk will discuss the genesis of the Foldit project and early results that pioneered the field of scientific discovery games for molecular biology. It will provide an overview of the foundation of this field of research.

Speakers:

Seth Cooper

02:20 PM - 02:40 PM (EDT)

Function - The Ortholog Conjecture Revisited: the Value of Orthologs and Paralogs in Function Prediction

The computational prediction of gene function is a key step in making full use of newly sequenced genomes. Function is generally predicted by transferring annotations from homologous genes or proteins for which experimental evidence exists. The ''ortholog conjecture'' proposes that orthologous genes should be preferred when making such predictions, as they evolve functions more slowly than paralogous genes. Previous research has provided little support for the ortholog conjecture, though the incomplete nature of the data cast doubt on the conclusions. Here we use experimental annotations from over 40,000 proteins, drawn from over 80,000 publications, to revisit the ortholog conjecture in two pairs of species: (i) Homo sapiens and Mus musculus and (ii) Saccharomyces cerevisiae and Schizosaccharomyces pombe. By making a distinction between questions about the evolution of function versus questions about the prediction of function, we find strong evidence against the ortholog conjecture in the context of function prediction, though questions about the evolution of function remain difficult to address. In both pairs of species, we quantify the amount of data that must be ignored if paralogs are discarded, as well as the resulting loss in prediction accuracy. Taken as a whole, our results support the view that the types of homologs used for function transfer are largely irrelevant to the task of function prediction. Aiming to maximize the amount of data used for this task, regardless of whether it comes from orthologs or paralogs, is most likely to lead to higher prediction accuracy.

Speakers:

Matthew Hahn

Indiana University

02:20 PM - 02:40 PM (EDT)

CAMDA - Spatial models for assessment of bacterial classification relevant to AMR

The work and development of the MetaSUB international consortium project raises important questions about the antimicrobial resistance (AMR) of the collected samples. Different bioinformatics and statistical approaches are developed for bacterial classification, AMR taxa discovery and characterization of their geographical distribution across multiple cities around the world. In this work we use methods from epidemiology to estimate relative risk of antimicrobial resistance by modeling the spatial correlation in the data. The novelty of our approach in this work is that we apply a convolution model or more specifically BYM (Besag, York, Mollié) model to incorporate explicitly the spatial structure in the data as determined by the longitude and latitude of the samples. We use the Bayesian setting implementation in R package CARBayes, where inference is based on Markov chain Monte Carlo (MCMC) simulation. We adapt the epidemiological concept of relative risk (RR) to find regions in the cities with elevated AMR risk. The model can also be used for data with excessive zeros by modeling the response as Zero Inflation Poisson (ZIP) process. In addition to spatial modeling, we used several machine learning approaches such as GBM, Random Forest and Neural Network to predict geographical origin of the samples.

Speakers:

Dimitar Vassilev

Sofia University "St. Kliment Ohridski", Faculty of Mathematics and Informatics

02:20 PM - 02:40 PM (EDT)

iRNA - Dissecting the role of SINE non-coding RNAs in amyloid pathology: An integrative RNA genomics approach.

As the human life span increases, the numbers of people in Canada suffering from aging associated cognitive impairment and Alzheimer’s disease (AD) are expected to rise dramatically. In a previous work we have shown that learning impairment is connected with epigenetic changes in stress response genes (SRGs),(Peleg*, Sanabenesi*,Zovoilis* et al, Science 2010). However, the molecular mechanisms associated with this epigenetic deregulation remain unknown. Among mechanisms that have recently attracted attention are those involving non-protein-coding RNAs (non-coding RNAs), including RNAs derived by repetitive DNA (Zovoilis et al, Cell 2016). Repetitive DNA accounts for ~50% of the noncoding sequences, with Short Interspersed Nuclear Elements (SINEs), being among the most frequent repeats. Here, we applied an integrative RNA genomics and bioinformatics approach to dissect any connection of SINE non-coding RNAs with amyloid pathology. Using short and long RNA sequencing, we demonstrate that SINE RNAs are associated with amyloid pathology in brain, revealing a potential biomarker and a novel molecular mechanism associated with this condition.

Speakers:

Athanasios Zovoilis

University of Lethbridge

02:25 PM - 02:30 PM (EDT)

Web - Overview 1

Speakers:

Javier De Las Rivas

Cancer Research Center (CiC-IBMCC, CSIC/USAL) & Global Organisation for Bioinformatics Learning, Education & Training (GOBLET)

02:30 PM - 02:35 PM (EDT)

Web - Overview 2

Speakers:

Bruno Gaeta

University of New South Wales

02:30 PM - 02:45 PM (EDT)

SST02 - Applying Citizen Science to Biomedical Literature Curation and Beyond

Biomedical literature is growing at a rate that outpaces our ability to harness the knowledge contained therein. In order to mine valuable inferences from the large volume of literature, many researchers have turned to information extraction algorithms to harvest information in biomedical texts. Advances in computational methods usually depends on the generation of gold standards by a limited number of expert curators. This process can be time consuming and represents an area of biomedical research that is ripe for exploration with citizen science. With our web-based application Mark2Cure (https://mark2cure.org), we demonstrate that citizen scientists can perform biocuration tasks on biomedical abstracts such as Named Entity Recognition (NER) and Relationship Extraction (RE). We discuss examples and challenges in applying citizen science to biomedical literature mining and explore additional biocuration and biomedical knowledge organization opportunities outside biomedical literature such as engaging citizen scientists to make biomedical research resources more findable, accessible, interoperable, and reusable.

Speakers:

Ginger Tsueng

Scripps Research

02:30 PM - 03:00 PM (EDT)

SST01 - The role of cis-regulatory elements in cell fate and immune disorders

Speakers:

David Hawkins

02:35 PM - 02:40 PM (EDT)

Web - Overview 3

Speakers:

Nicola Mulder

University of Cape Town

02:40 PM - 03:00 PM (EDT)

SysMod - Towards a Human Whole-Cell Model: A Prototype Model of Human Embryonic Stem Cells

Whole-cell (WC) models that integrate diverse data into a mechanistic understanding of every gene function could transform medicine and bioengineering. Towards this goal, we are prototyping a WC model of human embryonic stem cells, which have great potential for regenerative medicine. First, we combined multi-omics, biochemical, and physiological data into a structured knowledge base. Second, we used the knowledge base to generate submodels of multiple pathways, including metabolism, transcription, translation, macromolecular complexation, and RNA and protein degradation. Next, we integrated the submodels into a single model, and used the knowledge base to estimate the parameters and initial conditions of the model. We are using a multi-algorithmic approach to simulate the wide range of concentrations and fluxes involved in the model. We use stochastic simulation to simulate slow pathways such as transcription, we use flux balance analysis to simulate fast pathways such as metabolism, and we synchronize the shared variables of these sub-simulations throughout each simulation. To enable this work, we are developing several new databases, methods, and tools for building and simulating large models. Going forward, we aim to incorporate additional submodels of signal transduction and cell cycle regulation, and use the model to gain insights into stem cell self-renewal.

Speakers:

Yin Chew

Icahn School of Medicine at Mount Sinai

02:40 PM - 03:00 PM (EDT)

Text Mining - Comprehensive Named Entity Recognition on CORD-19 with Distant or Weak Supervision

Motivation: To facilitate biomedical text mining research on COVID-19, we have developed an automated, comprehensive named entity annotation and typing system CORD-NER. The system generates an annotated CORD-NER dataset based on the COVID-19 Open Research Dataset, covering 75 fine-grained types in high quality. Both the system and the annotated literature datasets will be updated timely. Results: The distinctive features of this CORD-NER dataset include (1) it covers 75 fine-grained entity types: In addition to the common biomedical entity types (e.g., genes, chemicals and diseases), it covers many new entity types specifically related to the COVID-19 studies (e.g., coronaviruses, viral proteins, evolution, materials, substrates and immune responses), which may benefit research on COVID-19 related virus, spreading mechanisms, and potential vaccines; (2) it relies on distantly- and weakly-supervised NER methods, with no need of expensive human annotation on any articles or sub-corpus; and (3) its entity annotation quality surpasses SciSpacy (over 10% higher on the F1 score based on a sample set of documents), a fully supervised BioNER tool. Our NER system supports incrementally adding new documents as well as adding new entity types when needed by adding dozens of seeds as the input examples.

Speakers:

Xuan Wang

University of Illinois at Urbana-Champaign

02:40 PM - 03:00 PM (EDT)

iRNA - Inferring competing endogenous RNA (ceRNA) interactions in cancer

To understand driving biological factors for cancer, regulatory circuity of genes needs to be discovered. Recently, a new gene regulation mechanism called competing endogenous RNA (ceRNA) interactions has been discovered. Certain RNAs targeted by common microRNAs (miRNAs) “compete” for these miRNAs, thereby regulate each other by making other free from miRNA regulation. Several computational tools have been published to infer ceRNA networks. In most existing tools, however, expression abundance and groupwise effect of ceRNAs are not considered. In this study, we developed a computational pipeline named Crinet to infer cancer-associated ceRNA networks addressing critical drawbacks. Crinet considers lncRNAs, pseudogenes and mRNAs as potential ceRNAs and incorporates a network deconvolution method to exclude amplifying effect of ceRNA pairs. We tested Crinet on breast cancer data in TCGA. Crinet inferred reproducible ceRNA interactions and groups, which were significantly enriched in cancer-related genes and biological processes. We validated our ceRNA interactions using protein expression data. Crinet outperformed existing tools predicting gene expression change in knockdown assays. Top high-degree genes in the inferred network included known suppressor/oncogene lncRNAs of breast cancer showing the importance of noncoding-RNA’s inclusion for ceRNA inference.

Speakers:

Ziynet Nesibe Kesimoglu

Marquette University

02:40 PM - 03:00 PM (EDT)

Function - Discovery of multi-operon colinear syntenic blocks in microbial genomes

Motivation: An important task in comparative genomics is to detect functional units by analyzing gene-context patterns. Colinear syntenic blocks (CSBs) are groups of genes that are consistently encoded in the same neighborhood and in the same order across a wide range of taxa. Such colinear syntenic blocks are likely essential for the regulation of gene expression in prokaryotes. Recent results indicate that colinearity can be conserved across multiple operons, thus motivating the discovery of multi-operon CSBs. This computational task raises scalability challenges in large datasets. Results: We propose an efficient algorithm for the discovery of cross-strand multi-operon CSBs in large genomic datasets. The proposed algorithm uses match-point arithmetic, which is scalable for large datasets of microbial genomes in terms of running time and space requirements. The algorithm is implemented and incorporated into a tool with a graphical user interface, denoted CSBFinder-S. We applied CSBFinder-S to data mine 1,485 prokaryotic genomes and analyzed the identified cross-strand CSBs. Our results indicate that most of the syntenic blocks are exclusively colinear. Additional results indicate that transcriptional regulation by overlapping transcriptional genes is abundant in bacteria. We demonstrate the utility of CSBFinder-S to identify common function of the gene-pair PulEF in multiple contexts, including Type 2 Secretion System, Type 4 Pilus System, and DNA uptake.

Speakers:

Dina Svetlitsky

Ben Gurion University of the Negev

02:40 PM - 03:00 PM (EDT)

CAMDA - Metagenomic Geolocation Prediction Using an Adaptive Ensemble Classifier

Several standard classifiers have been employed for the prediction of the origin of a given microbial sample. The application of these individual classification algorithms yields a varying degree of classification accuracy, and the performance of such classifiers are also dependent on the structure of the available data. Rather than employing several individual classifiers, we adopt the adaptive ensemble classification algorithm proposed by (Datta, 2010). The ensemble classifier, which is constructed by bagging and rank aggregation, comprises a set of standard classification algorithms where such individual algorithms are combined flexibly to yield classification performance as least, as good as the best classification algorithm in the ensemble. For our analysis, we trained and tested fourteen standard classifiers including the ensemble classifier, and in different instances, we also applied class weighting and an optimal oversampling technique to overcome the problem of class balance in the primary data. These analyses were conducted both on the primary data set of relative abundances and data with feature reduced space. In each instance, we found that the standard algorithms performed differently, whereas the ensemble classifier consistently showed to have optimal performance. Lastly, we predict the source cities of the mystery samples provided by the CAMDA organizers.

Speakers:

Samuel Anyaso-Samuel

University of Florida

02:40 PM - 03:00 PM (EDT)

Web - Group discussion on University specific aspects of online training: Resources in place, numbers taught, assessment methods

Speakers:

Bruno Gaeta

University of New South Wales

Javier De Las Rivas

Cancer Research Center (CiC-IBMCC, CSIC/USAL) & Global Organisation for Bioinformatics Learning, Education & Training (GOBLET)

Nicola Mulder

University of Cape Town

Tugce Bilgin

Columbia University

02:45 PM - 03:00 PM (EDT)

SST02 - Phylo: How to turn scientific tasks into casual games

Launched in 2010, Phylo is a casual tile-matching game aiming to improve large-scale multiple sequence alignments. In this talk, we will discuss the origins and evolution of this scientific game during the last decade. We will discuss its motivations, its impact and various lessons we learned through this project.

Speakers:

Roman Sarrazin-Gendron

McGill University

03:19 PM - 03:20 PM (EDT)

Web - RECENT ONLINE TRAINING EXPERIENCES - SHORT-COURSE EXPERIENCE

Speakers:

Cath Brooksbank

03:20 PM - 03:30 PM (EDT)

SST01 - Uncovering B-ALL TF-gene regulatory interactions associated with CRLF2-overexpression

Although genetic alterations are initial drivers of disease, aberrantly activated transcriptional regulatory programs are often responsible for maintenance and progression in cancer. Alterations leading to CRLF2-overexpression in B-ALL patients are associated with poor outcome and activate JAK-STAT, PI3K and ERK/MAPK signaling pathways. Although inhibitors of these pathways are available, there still remains the issue of treatment-associated toxicities and poorly studied regulatory structures controlling leukemogenesis. Comparing RNA-seq from CRLF2-High and other B-ALL patients (CRLF2-Low), we defined a CRLF2-High gene signature. Patient-specific chromatin accessibility was interrogated to identify altered putative regulatory elements that could be linked to transcriptional changes. To delineate these regulatory interactions, a B-ALL cancer-specific regulatory network was inferred using 868 B-ALL patient samples from the NCI TARGET database coupled with priors generated from ATAC-seq peak TF-motif analysis. Analysis of CRISPRi, siRNA knockdown and ChIP-seq of nine TFs involved in the inferred network were analyzed to validate a cohort of predicted TF-gene regulatory interactions. Inferred interactions were used to identify differential patient-specific transcription factor activities predicted to control CRLF2-High deregulated genes, thereby enabling identification of robust gene targets.

Speakers:

Sana Badri

Department of Pathology, New York University School of Medicine, New York, NY, USA.

03:20 PM - 03:40 PM (EDT)

SysMod - Whole-body regeneration and size-dependent fission controlled by a self-regulated Turing system in planaria

Planarian worms have the extraordinary ability to regenerate any body part after an amputation. This ability allows them to reproduce asexually by fission, cutting themselves to produce two separated pieces each repatterning and regenerating a complete animal. The induction of this process is known to be dependent on the size of the worm as well as on environmental factors such as population density, temperature, and light intensity. Models based on Turing systems can explain the self-regulation of many biological mechanisms, from skin patterns to digit formation. Here, we combine experimental evidence with a modeling approach to show how a cross-inhibited Turing system can explain at once both the signaling mechanism of regeneration and fission in planaria. The model explains in a growing domain the precise signals that control the regeneration of the different body parts after amputations as well as when and where planaria fission, and its dependence on the worm length. We provide molecular implementations of the proposed model, which also explains the effects of environmental factors in the signaling of fission. In summary, the proposed controlled cross-inhibited Turing system represents a completely self-regulated model of the whole-body regeneration and fission signaling in planaria.

Speakers:

Daniel Lobo

University of Maryland, Baltimore County

03:20 PM - 03:25 PM (EDT)

Web - Overview 4

Speakers:

Patricia Palagi

SIB Swiss Institute of Bioinformatics

03:20 PM - 03:35 PM (EDT)

SST02 - Improved RNA secondary structure modeling through crowdsourced RNA design initiatives

Although algorithms for predicting RNA structure have been in development for decades, little is known about their accuracy in many of their use-cases. We created a database, "EternaBench", comprising the diverse high-throughput structural data gathered through the crowdsourced RNA design project Eterna, to evaluate the performance of a wide set of structure algorithms. Surprisingly, we found that lesser-used algorithms, which were developed through statistical learning, perform notably better than widely-used algorithms that are derived from experimental data. Motivated by this finding, we developed a multitask-learning-based model, "EternaFold", trained on the EternaBench data, which demonstrates improved predictions both on molecules from Eterna, as well as completely independent datasets of viral genomes and mRNAs. Experimental data from the project Eterna allows us to establish the first large-scale, independent benchmarks of RNA thermodynamic predictions, as well as leverage the diversity of player designs to improve structure prediction through statistical learning.

Speakers:

BrombergLab, Rutgers University

03:20 PM - 04:00 PM (EDT)

Function - Function COSI Keynote: Saving Time at the Bench and in the Field: Predicting Gene Function and Phenotype in Crops

Speakers:

Carolyn Lawrence-Dill

Iowa State University

03:25 PM - 03:30 PM (EDT)

Web - Overview 5

Speakers:

Sarah Morgan

EMBL-EBI

03:30 PM - 03:35 PM (EDT)

Web - Overview 6

Speakers:

Rachade Hmamouchi

CBW and MiCM

03:30 PM - 03:40 PM (EDT)

Text Mining - COVID-19 systems demo panel (part 2)

Speakers:

Jiawei Han

University of Illinois

03:30 PM - 04:00 PM (EDT)

SST01 - Examining waning immunity of B. pertussis vaccination

Speakers:

Bjoern Peters

La Jolla Institute for Immunology

03:35 PM - 03:50 PM (EDT)

SST02 - Stall Catchers: Engineering Obsolescence

It might seem ironic that the goal of a citizen science project would be to render itself obsolete, but it turns out this could be the most fruitful approach for science and the most ethical approach for online participants. Stall Catchers is an online game created to accelerate Alzheimer's research at Cornell University. Today, over 25,000 registered Stall Catchers volunteers are analyzing datasets in 1 or 2 months that formerly required up to a year to analyze in the lab. Continuing to expand the participant community could further accelerate this analysis, however, a more efficient path might entail using the human generated data to improve machine-based contributions. This talk describes how the evolution of the Stall Catchers project may suggest a life cycle for future generations of citizen science projects that respects their contributors while maximizing societal impacts.

Speakers:

Pietro Michelucci

Human Computation Institute

03:35 PM - 03:53 PM (EDT)

Web - Group discussion on short-course specific aspects of online training: Technologies used, hands-on facilitation, numbers taught

Speakers:

Rachade Hmamouchi

CBW and MiCM

Sarah Morgan

EMBL-EBI

Patricia Palagi

SIB Swiss Institute of Bioinformatics

03:40 PM - 03:50 PM (EDT)

Text Mining - COVID-19 systems demo panel (part 3)

Speakers:

Tonia Korves

MITRE

03:40 PM - 04:00 PM (EDT)

MLCSB: Machine Learning - Dissecting the grammar of chromatin architecture using megabase scale DNA sequence with deep neural networks and transfer learning

Mammalian genome architecture is characterized by an intricate framework of hierarchically folded domains that dictate essential gene regulatory functions such as enhancer-promoter interactions. Understanding which determinants are driving 3D genome formation and how this architecture is altered through the sequence and structural variations requires high-throughput, genome-wide approaches. Computational models sophisticated enough to grasp the determinants of chromatin folding allow us to perform these large scale experiments in silico. We have developed a deep neural network (deepC) that uses transfer learning to predict chromatin interactions from DNA sequence at the megabase scale. Our model predicts Hi-C interactions at high resolution captures intricate, hierarchical chromatin structures and can be used to fine-map Hi-C data. DeepC allows us to predict the impact of single base pair variants as well as structural variation in the same end-to-end framework, bridging the different levels of resolution from base pairs to TADs. DeepC enables large-scale, computational screens that empower us to dissect the functional elements and sequence determinants that regulate chromatin architecture at base pair resolution and genome-wide scale. We demonstrate how we employ deepC to stratify the contribution of distinct classes of regulatory elements, study the grammar of domain boundaries and predict the effect of SNPs.

Speakers:

Ron Schwessinger

University of Oxford

03:40 PM - 04:00 PM (EDT)

SysMod - Clb3-centered regulations are pivotal for autonomous cell cycle oscillator designs in yeast

Some biological networks exhibit oscillations in their components to convert stimuli to time-dependent responses. The eukaryotic cell cycle is such a case, being governed by waves of cyclin-dependent kinase (cyclin/Cdk) activities that rise and fall with specific timing and guarantee its timely occurrence. Disruption of cyclin/Cdk oscillations could result in dysfunction through reduced cell division. Therefore, it is of interest to capture properties of network designs that exhibit robust oscillations. Here we show that a minimal cell cycle network in budding yeast is able to oscillate autonomously, and that cyclin/Cdk-mediated positive feedback loops (PFLs) and Clb3-centered regulations sustain cyclin/Cdk oscillations, in known and hypothetical network designs. An integrative, computational and experimental approach pinpoints how robustness of cell cycle control is realized by revealing a novel and conserved principle of design that ensures a timely interlock of transcriptional and cyclin/Cdk oscillations. Given the evolutionary conservation of the cell cycle network across eukaryotes, the cyclin/Cdk network can be used as a core building block of multi-scale models that integrate regulatory modules to address cellular physiology.

Speakers:

Matteo Barberis

University of Surrey

03:40 PM - 04:00 PM (EDT)

iRNA - A Bayesian framework for inter-cellular information sharing improves dscRNA-seq quantification

Motivation: Droplet based single cell RNA-seq (dscRNA-seq) data is being generated at an unprecedented pace, and the accurate estimation of gene level abundances for each cell is a crucial first step in most dscRNA-seq analyses. When preprocessing the raw dscRNA-seq data to generate a count matrix, care must be taken to account for the potentially large number of multi-mapping locations per read. The sparsity of dscNRA-seq data, and the strong 3’ sampling bias, makes it difficult to disambiguate cases where there is no uniquely mapping read to any of the candidate target genes. Results: We introduce a Bayesian framework for information sharing across cells within a sample, or across multiple modalities of data using the same sample, to improve gene quantification estimates for dscRNA-seq data. We use an anchor-based approach to connect cells with similar gene expression patterns, and learn informative, empirical priors which we provide to alevin’s gene multi-mapping resolution algorithm. This improves the quantification estimates for genes with no uniquely mapping reads (i.e. when there is no unique intra-cellular information). We show our new model improves the per cell gene level estimates and provides a principled framework for information sharing across multiple modalities. We test our method on a combination of simulated and real datasets under various setups. Availability: The information sharing model is included in alevin and is implemented in C++14. It is available as open-source software, under GPL v3, at https://github.com/COMBINE-lab/salmon as of version 1.1.0.

Speakers:

Avi Srivastava

New York Genome Center.

03:40 PM - 04:00 PM (EDT)

CAMDA - Metagenomic Data Analysis with Probability-Based Reduced Dataset Representation

Metagenomics has become of increasing interest to researchers exploring human health and disease. Higher sequencing throughput and lower associated costs, in turn, has made whole genome sequencing of complex environmental samples a realistic possibility for researchers. The massive datasets produced by these approaches, however, calls for new approaches to their processing and interpretation. MetaSUB and CAMDA have partnered together to confront this challenge, offering a yearly contest in which participants are asked to predict the origin of unlabeled metagenomic samples by applying machine learning methods to a set of city-labeled whole genome sequencing data. A remaining roadblock, however, is the enormous computational power required to process these massive datasets. Previous research has studied the effect of mystery sample prediction using a "reduced-representation subset" which contained only 50% of the sample data in the training set. In each case, there is relatively little loss of predictive power using the subset. In this paper, we extend upon the previous work and explore the relationship between mystery sample prediction and reduced-representation subsets at 20% intervals.

Speakers:

Cory Gardner

Saint Louis University

03:50 PM - 04:05 PM (EDT)

SST02 - Accelerating Microbiome Science by Empowering Individuals

The human microbiome plays a fundamental role in human health. These microbes, that live on and inside of us, perform a range of critical functions, including vitamin synthesis and the production of short chain fatty acids. Research over the past decade has highlighted relationships between the microbiome and diseases like Inflammatory Bowel Disease, Nonalcoholic Fatty Liver Disease, Parkinson’s, cancer and more. Concurrently, research has shown a remarkable degree of variability in the microbiome among individuals, with substantially more variation than peoples genomes. The factors related to the dynamic ranges of microbes observed are not well understood, nor is it known whether certain configurations are healthier or may predispose you to disease. Two major limitations are data: a small portion of the human population has been sampled, and the collective set of microbiome studies performed have incompatible standards. Within the American Gut Project, and its global extension The Microsetta Initiative, we are combating these limitations through infrastructure that allows virtually anyone to participate. To date, over 20,000 have submitted microbiome samples. The data produced, including a detailed voluntary questionnaire, are released de-identified into the public domain for anyone to reuse. In this talk, we discuss the importance of citizen science for the microbiome, and some of the challenges encountered since project inception in the fall of 2012.

Speakers:

Daniel McDonald

03:50 PM - 04:00 PM (EDT)

Text Mining - COVID-19 systems demo panel (part 4)

Speakers:

Fabio Rinaldi

IDSIA

03:53 PM - 03:55 PM (EDT)

Web - BEST PRACTICES IN ONLINE TRAINING

Speakers:

Nicola Mulder

University of Cape Town

03:55 PM - 04:10 PM (EDT)

Web - Best practices from the experience: the Carpentries

Speakers:

Kate Hertweck

Fred Hutchinson Cancer Research Center

04:00 PM - 04:40 PM (EDT)

CAMDA-Towards a metagenomics interpretable model for understanding the transition from adenoma to colorectal cancer

TBP

Speakers:

Carlos Loucera

Clinical Bioinformatics Area (FPS)

04:00 PM - 04:10 PM (EDT)

Text Mining - COVID-19 systems demo panel (part 5)

Speakers:

John Bachman

Laboratory of Systems Pharmacology, Harvard Medical School

04:00 PM - 04:20 PM (EDT)

MLCSB: Machine Learning - CoRE-ATAC: A Deep Learning model for the Classification of Regulatory Elements from single cell and bulk ATAC-seq data

Cis-Regulatory elements (cis-REs) include promoters, enhancers, and insulators that regulate gene expression programs via binding of transcription factors. ATAC-seq technology effectively identifies active cis-REs in a given cell type (including from single cells) by mapping the accessible chromatin at base-pair resolution. However, these maps are not immediately useful for inferring specific functions of cis-REs. For this purpose, we developed a deep learning framework (CoRE-ATAC) with novel data encoders that integrate DNA sequence (reference or personal genotypes) and ATAC-seq read pileups. CoRE-ATAC was trained on 4 cell types (n=6 samples/replicates) and accurately predicted known cis-RE functions from 7 cell types (n=40 samples) that were not used in model training (average precision=0.80). CoRE-ATAC enhancer predictions from 19 human islets coincided with genetically modulated gain/loss of enhancer activity, which was confirmed by massively parallel reporter assays (MPRAs). Finally, CoRE-ATAC effectively inferred functionality of cis-REs from single nucleus ATAC-seq data from human blood-derived immune cells that overlapped well with known functional annotations in sorted immune cells. ATAC-seq maps from primary human cells reveal individual- and cell-specific variation in cis-RE activity. CoRE-ATAC increases the functional resolution of these maps, a critical step for studying regulatory disruptions behind diseases.

Speakers:

Asa Thibodeau

The Jackson Laboratory for Genomic Medicine, Farmington, CT, 06032, USA.

04:00 PM - 04:20 PM (EDT)

Function - DeepPheno: Predicting single gene loss-of-function phenotypes using an ontology-aware hierarchical classifier

Predicting the phenotypes resulting from molecular perturbations is one of the key challenges in genetics. Both forward and reverse genetic screen are employed to identify the molecular mechanisms underlying phenotypes and disease, and these resulted in a large number of genotype--phenotype association being available for humans and model organisms. Combined with recent advances in machine learning, it may now be possible to predict human phenotypes resulting from particular molecular aberrations. We developed DeepPheno, a neural network based hierarchical multi-class multi-label classification method for predicting the phenotypes resulting from complete loss-of-function in single genes. DeepPheno uses the functional annotations with gene products to predict the phenotypes resulting from a loss-of-function; additionally, we employ a two-step procedure in which we predict these functions first and then predict phenotypes. Prediction of phenotypes is ontology-based and we propose a novel ontology-based classifier suitable for very large hierarchical classification tasks. These methods allow us to predict phenotypes associated with any known protein-coding gene. We evaluate our approach using evaluation metrics established by the CAFA challenge and compare with top performing CAFA2 methods as well as several state of the art phenotype prediction approaches, demonstrating the improvement of DeepPheno over state of the art methods.

Speakers:

Maxat Kulmanov

King Abdullah University of Science and Technology

04:00 PM - 04:20 PM (EDT)

SysMod - Robust Inference of Kinase Activity Using Functional Networks

Recent developments in mass spectrometry (MS) enable high-throughput screening of phospho-proteins across a broad range of biological contexts. Phospho-proteomic data complemented by computational algorithms enable the inference of kinase activity facilitating the identification of dysregulated kinases in various diseases, including cancer, Alzheimer’s disease and Parkinson’s disease, among others. However, the inadequacy of known kinase-substrate associations and the incompleteness of MS-based phosphorylation data pose important limitations on the inference of kinase activity. With a view to enhancing the reliability of kinase activity inference, we present a network-based framework named RoKAI that integrates various sources of functional information. These functional information include structure distance, co-evolution evidence, shared kinase associations, and protein-protein interaction networks. By propagating phosphorylation data across these networks, RoKAI obtains representative phosphorylation profiles capturing coordinated changes in signaling. The resulting phosphorylation profiles can be used in conjunction with any existing or future inference methods to predict kinase activity. The results of our computational experiments show that RoKAI consistently improves the accuracy of commonly used kinase activity inference methods and makes them more robust to missing kinase-substrate annotations. To provide an easy to use interface to users, RoKAI is available as a web-based tool at http://rokai.ngrok.io.

Speakers:

Serhan Yılmaz

Case Western Reserve University

04:00 PM - 04:20 PM (EDT)

SST01 - Sexual dimorphism in human immune system aging

Speakers:

Duygu Ucar

The Jackson Laboratory for Genomic Medicine, Farmington, CT, 06032, USA.

04:00 PM - 04:40 PM (EDT)

iRNA - iRNA Keynote: The race to the 3’ end: tracking mRNA cleavage diversity in real time

Alternative mRNA processing is increasingly appreciated to play a large role in driving transcriptome variability, disease etiology, cellular identity, and the molecular response to diverse environmental stresses. Although we have extensive insights into the regulatory factors and sequence elements that influence alternative isoform usage, less is known about the temporal dynamics and co-regulation of RNA processing decisions. Intermediate processing events—splicing and 3’ end cleavage—often occur co-transcriptionally, with the interplay between transcriptional elongation rates and rates of each processing event often impacting choices that lead to alternative isoform production. Consequently, measuring the kinetics of these processes may shed light on how early gene regulatory decisions are made. Recent development of high-throughput sequencing techniques that capture nascent RNA over defined temporal intervals has made genome-wide kinetic profiling of RNA maturation possible. Though rates of mRNA splicing have been estimated globally, the rate at which an mRNA is cleaved and polyadenylated to complete the maturation process has never been investigated. Here, we present a novel computational method to estimate genome-wide kinetic parameters for mRNA cleavage rates. This method capitalizes on short-read sequencing data from nascent mRNAs isolated after a time-course of 4sU metabolic labeling to model the rate of mRNA maturation over time. To specifically measure cleavage rates, we first use patterns of read coverage from our sequencing data to approximate the position at which cleavage occurs. We then estimate the fraction of reads derived from cleaved or uncleaved molecules at that site across time to model the rate of 3’ end cleavage over time. We applied this method to nascent RNA-seq data from Drosophila melanogaster S2 cells to estimate polyadenylation-site (PAS) specific rates of mRNA cleavage. Our findings shed light on the timing of decisions involved in alternative PAS usage within genes and the variable efficiency of 3’ end cleavage and polyadenylation across genes.

Speakers:

Athma Pai

University of Massachusetts Medical School

04:05 PM - 04:20 PM (EDT)

SST02 - Gamers and experimentalists collaborate on COVID-19

Scientific games allow users to process raw experimental data and submit hypothesis to scientists that can be validated in labs. This theme will explore opportunities arising with the creation of new bridges between citizens and scientists, highlighting recent successes with the Foldit project. Citizen scientists have collaborated with scientists in an attempt to design anti-viral and anti-inflammatories against COVID-19.

Speakers:

Firas Khatib

University of Massachusetts, Dartmouth

04:10 PM - 04:25 PM (EDT)

Web - Best practices from the experience: Software Sustainability

Speakers:

Rachael Ainsworth

Software Sustainability Institute

04:10 PM - 04:40 PM (EDT)

Text Mining - COVID-19 systems demo panel (part 6)

Speakers:

John Bachman

Laboratory of Systems Pharmacology, Harvard Medical School

Fabio Rinaldi

IDSIA

Tonia Korves

MITRE

Jiawei Han

University of Illinois

Xuan Wang

University of Illinois at Urbana-Champaign

Amalie Trewartha

Lawrence Berkeley National Laboratory

04:20 PM - 04:30 PM (EDT)

SST01 - Integrated transcriptomic analysis of SLE reveals IFN-driven cross-talk between immune cells

Systemic lupus erythematosus (SLE) is an incurable autoimmune disease disproportionately affecting women and may lead to damage in multiple different organs. The marked heterogeneity in its clinical manifestations is a major obstacle in finding targeted treatments and involvement of multiple immune cell types further increases this complexity. Thus, identifying molecular sub-types that best correlate with disease heterogeneity and severity as well as deducing molecular cross-talk among major immune cell types that lead to disease progression are critical steps in development of more informed therapies for SLE. Here we profile and analyze gene expression of six major circulating immune cell types from patients with well-characterized SLE and from healthy subjects. Our results show that the interferon (IFN) response signature was the major molecular feature that classified SLE patients into two distinct groups. We show that the gene expression signature of IFN response was consistent among different datasets and cell types. For a better understanding of molecular differences of IFNpos versus IFNneg patients, we combined differential gene expression analysis with differential Weighted Gene Co-expression Network Analysis (WGCNA), which revealed a relatively small list of genes. We have also performed the integrated WGNCA (iWGCNA) to understand the cross-talk between different immune cell types.

Speakers:

Bharat Panwar

La Jolla Institute for Immunology

04:20 PM - 04:35 PM (EDT)

SST02 - Crowdsourced design of stabilized COVID-19 mRNA vaccines with Eterna OpenVaccine

While more rapidly deployable than other kinds of vaccines, COVID-19 mRNA vaccines have a problem: their poor chemical stability at refrigerator temperatures precludes worldwide delivery in prefilled syringes, as would be needed for immediate mass immunization. Recent studies have implicated mRNA structure as a critical factor in stability and expression, but our ability to rationally design these desired properties remains primitive. In March 2020, we launched the OpenVaccine challenge on Eterna (https://eternagame.org) to accelerate the design of stabilized, high-expression mRNA vaccines through a worldwide, open competition. The project involves an internet-scale videogame with mRNA design challenges; novel methods for rapid synthesis of 100s to 1000s of long mRNAs per round and for assessment of mRNA degradation and protein expression in human cells; and an IP framework ensuring open use of results while allowing companies to license exclusive redesign of their mRNA vaccine candidates.

Speakers:

Rhiju Das

Stanford University

04:20 PM - 04:40 PM (EDT)

MLCSB: Machine Learning - Zero-shot imputations across species are enabled through joint modeling of human and mouse epigenomics

Recent large-scale efforts to characterize biochemical activity along the human genome have produced thousands of genome-wide experiments that quantify various forms of biological activity, such as histone modifications, protein binding, and chromatin accessibility. Although these experiments represent a small fraction of the possible experiments that could be performed, the human genome remains the most characterized of any species. We propose an extension to the imputation approach Avocado that enables the model to leverage the large number of human genomic data sets when making imputations in other species. This extension takes advantage of shared synteny and the similar role that many types of biochemical activity have in evolutionarily-related species. We found that not only does this extension result in improved imputations of mouse genomics experiments, but that the extended model is able to make accurate imputations for assays that have been performed in humans but not in mice. This ability to make ``zero-shot'' imputations greatly increases the utility of such imputation approaches, and enables comprehensive imputations to be made for species even when experimental data are sparse.

Speakers:

Jacob Schreiber

Paul G. Allen School of Computer Science and Engineering, University of Washington, Seattle, USA

04:20 PM - 04:40 PM (EDT)

Function - HPOLabeler: Improving Prediction of Human Protein-Phenotype Associations by Learning to Rank

Annotating human proteins by abnormal phenotypes has become an important topic. As of Nov. 2019, only less than 4,000 proteins have been annotated with Human Phenotype Ontology (HPO). Thus a computational approach for accurately predicting protein-HPO associations would be important, while no methods have outperformed a simple Naive approach in the CAFA2 (second Critical Assessment of Functional Annotation, 2013-14). We present HPOLabeler, which can use a wide variety of evidence, such as protein-protein interaction networks (PPI), Gene Ontology (GO), InterPro, trigram frequency and HPO term frequency, in the framework of learning to rank (LTR). Given an input protein, LTR outputs the ranked list of HPO terms from a series of input scores given to the candidate HPO terms by component learning models (logistic regression, nearest neighbor and a Naive method), which are trained from given multiple evidence. We empirically evaluate HPOLabeler extensively through mainly two experiments of cross-validation and temporal validation, for which HPOLabeler significantly outperformed all component models and competing methods including the current state-of-the-art method. We further found that 1) PPI is most informative for prediction among diverse data sources, and 2) low prediction performance of temporal validation might be caused by incomplete annotation of new proteins.

Speakers:

Lizhi Liu

Fudan University

04:20 PM - 04:40 PM (EDT)

SysMod - Probabilistic Factor Graph Modeling and Analysis of Biological Networks

Reverse engineering of molecular networks from biological data is one of the most challenging tasks in systems biology. Numerous inference and computational methodologies have been formalized to enhance the deduction of reliable and testable predictions in today’s biology. However, there is little research aimed to quantify how well the existing state-of-the-art molecular networks correspond to the measured gene expressions. We present a computational framework that combines formulation of probabilistic graphical model, standard statistical estimation, and integration of high-throughput gene expression data. The model is represented as a probabilistic bipartite graph, which accommodates partial information of diverse biological entities to study and analyze the global behavior of a biological system. This result provides a building block for performing simulations on the consistency between inferred gene regulatory networks and corresponding biological data. We test the applicability of our model to explore the allowable stable-states in two experimentally verified regulatory pathways in Escherechia Coli using real microarray expression data from the M3D database. Furthermore, the model is employed to quantify how well the pathways are explained by the extant microarray data. Results show a surprisingly high correlation between the observed states of the experimental data under various conditions and the inferred system’s behavior.

Speakers:

Stephen Kotiang

Wichita State University

04:25 PM - 04:40 PM (EDT)

Web - Education Summit report

Speakers:

Venkata Satagopam

University of Luxembourg

04:30 PM - 04:40 PM (EDT)

SST01 - Semi-connected multilayer perceptron for single-cell profiling

Speakers:

Grégoire Altan-Bonnet

National Cancer Institute

James Anibal

National Cancer Institute

04:40 PM - 04:50 PM (EDT)

MLCSB: Machine Learning - A Combined Species Model using Branched Multitask Routing Networks

Recent development in deep learning has shown unprecedented accuracy in predicting chromatin features using DNA sequences alone. The high prediction accuracy partly stems from the millions of labeled sequences available across whole organs and cell-lines in culture, but for specific biological systems, the number of labeled sequences is often an order of magnitude lower. Acquiring additional labeled sequences from the same species adds little variability. A promising strategy is to combine datasets across species, since flanks around conserved regulatory subsequences tend to vary. Common approaches for combining datasets include transfer learning and multitask learning. However, these approaches do not facilitate common and species-specific attribute extraction from the models. To improve both prediction accuracy and model interpretability, we propose a branched multitask routing network (BMTRN) for cross species chromatin feature prediction. The idea is to split a network into common and species-specific layers via task routing so that shared signals between datasets can be exploited without assuming the species have the same regulatory attributes. We apply BMTRN to ATAC-seq datasets from mouse and human, and show that BMTRN improves prediction accuracy, enhances filter reproducibility, facilitates easier interpretation of cross-species differences, and increases sensitivity in detecting effects of functional variants in silico.

Speakers:

Chendi Wang

The University of British Columbia

05:00 PM - 05:20 PM (EDT)

Function - Enzymes, Moonlighting Enzymes, Pseudoenzymes: Similar in Sequence, Different in Function

The function of a newly sequenced protein is often estimated by sequence alignment with the sequences of proteins with known functions. However, members of a protein superfamily can share significant amino acid sequence identity but vary in the reaction catalyzed and/or the substrate used. In addition, a protein superfamily can include moonlighting proteins, which have two or more functions, and pseudoenzymes, which have a three-dimensional fold that resembles a conventional catalytically active enzyme, but do not have catalytic activity. I will discuss several examples of protein families that contain enzymes with noncanonical catalytic functions, pseudoenzymes, and/or moonlighting proteins. Pseudoenzymes and moonlighting proteins are widespread in the evolutionary tree and are found in many protein families, and they are often very similar in sequence and structure to their monofunctional and catalytically active counterparts. A greater understanding is needed to clarify when similarities and differences in amino acid sequences and structures correspond to similarities and differences in biochemical functions and cellular roles. This information can help improve programs that identify protein functions from sequence or structure and assist in more accurate annotation of sequence and structural databases, as well as in our understanding of the broad diversity of protein functions.

Speakers:

Constance Jeffery

University of Illinois at Chicago

05:00 PM - 05:20 PM (EDT)

iRNA - RBP-Pokedex: Prediction of RBP knockdown effect via DNN experiment modeling

The genome of eukaryotes code for hundreds of RNA-binding proteins (RBP), which regulate the fate of RNA from synthesis to degradation. These RBPs form an extensive network of condition specific regulation, including direct and indirect regulation of other RBPs. This fact, combined with technological limitations, make systematic experimental characterization of their effect on gene expression infeasible. Here, we propose to leverage the available expression measurements in diverse conditions to instead build predictive models for the effect of RBP knockdown on expression of other RBPs, capturing a condition-specific “RBP state”. We develop a model for prediction of knockdown effects via DNN experiments (Pokedex). We use an unsupervised learning approach where we construct a variational autoencoder for the expression of 531 RBPs. Training it on the naturally occurring variations of RBP expression across 53 GTEx tissues we test its ability to predict knockdown effects in ENCODE experiments. Using the expression changes of RBPs compared to control as the test statistic we show this DNN performs significantly better then a standard PCA based linear model of variation. Finally, we make the learned model available to the RNA community through a web-tool, RBP-Pokedex (https://tools.biociphers.org/rbp-pokedex), to predict the effect of pan-tissue RBP knockdowns.

Speakers:

Anupama Jha

University of Pennsylvania

05:00 PM - 05:20 PM (EDT)

SysMod - Executable models of pathways built using single-cell RNAseq data reveal immune cell heterogeneity in people living with HIV and atherosclerosis

Single-cell RNA-sequencing (scRNA-seq) enables the profiling of mixed tissues at unprecedented resolution. Clustering of these sequenced cells into cell types is a major challenge, compounded by cell heterogeneity in terms of signaling pathways that cause observed phenotypes. Knowledge-driven methods utilize cell type specific marker genes for clustering. However, these methods do not account for the pathway-level transcriptional differences that define cell types. We propose a knowledge-driven clustering method using Boolean modeling of transcriptional networks. We have previously developed an algorithm called BONITA that infers Boolean rules characterizing pre-defined topologies from bulk transcriptomic data [1]. We demonstrate the utility of our approach on an scRNA-seq dataset of peripheral blood mononuclear cells from HIV+ individuals with and without atherosclerosis. Specifically, we (i) simulated networks and identified reachable attractors (ii) identified subpopulation-specific pathways dysregulated in HIV-associated atherosclerosis and (iii) clustered cells in multi-attractor space. We propose that cell clusters such as CD14+ CD16+ monocytes, that are implicated in HIV-associated atherosclerosis, can be characterized based on pathway-level similarities obtained using Boolean pathway models. References: 1. Palli R, Palshikar MG, Thakar J (2019) Executable pathway analysis using ensemble discrete-state modeling for large-scale data. PLoS Comput Biol 15(9): e1007317.

Speakers:

Mukta Palshikar

University of Rochester

05:00 PM - 05:20 PM (EDT)

Text Mining - PEDL: Extracting protein-protein associations using deep language models and distant supervision.

Motivation: A significant portion of molecular biology investigates signalling pathways and thus depends on an up-to-date and complete resource of functional protein-protein associations (PPAs) that constitute such pathways. Despite extensive curation efforts, major pathway databases are still notoriously incomplete. Relation extraction can help to gather such pathway information from biomedical publications. Current methods for extracting PPAs typically rely exclusively on rare manually labelled data which severely limits their performance. Results: We propose PEDL, a method for predicting PPAs from text that combines deep language models and distant supervision. Due to the reliance on distant supervision, PEDL has access to an order of magnitude more training data than methods solely relying on manually labelled annotations. We introduce three different data sets for PPA prediction and evaluate PEDL for the two subtasks of predicting PPAs between two proteins, as well as identifying the text spans stating the PPA. We compared PEDL with a recently published state-of-the-art model and found that on average PEDL performs better in both tasks on all three data sets. An expert evaluation demonstrates that PEDL can be used to predict PPAs that are missing from major pathway data bases and that it correctly identifies the text spans supporting the PPA. Availability: PEDL is freely available at https://github.com/leonweber/pedl. The repository also includes scripts to generate the used data sets and to reproduce the experiments from this paper. Contact: [email protected] or [email protected] Supplementary information: Supplementary data are available at Bioinformatics online.

Speakers:

Leon Weber

Humboldt-Universität zu Berlin

05:00 PM - 05:20 PM (EDT)

CAMDA - IDMIL: An alignment-free interpretable deep multiple instance learning (MIL) for predicting disease from whole-metagenomic data

The human body hosts more microbial organisms than human cells. Analysis of this microbial diversity provides key insight into the role played by these microorganisms on human health. Metagenomics is the collective DNA sequencing of coexisting microbial organisms in a host. This has several applications in precision medicine, agriculture, environmental science, and forensics. State-of-the-art predictive models for phenotype predictions from microbiome rely on alignments, assembly, extensive pruning, taxonomic profiling along with expert-curated reference databases. These processes are time-consuming and they discard the majority of the DNA sequences for downstream analysis limiting the potential of whole-metagenomics. We formulate the problem of predicting human disease from whole-metagenomic data using Multiple Instance Learning (MIL), a popular supervised learning paradigm. Our proposed alignment-free approach provides higher accuracy in prediction by harnessing the capability of deep convolutional neural network (CNN) within a MIL framework and provides interpretability via neural attention mechanism. The MIL formulation combined with the hierarchical feature extraction capability of deep-CNN provides significantly better predictive performance compared to popular existing approaches. The attention mechanism allows for the identification of groups of sequences that are likely to be correlated to diseases providing the much-needed interpretation. Our proposed approach does not rely on alignment, assembly and manually curated databases; making it fast and scalable for large-scale metagenomic data. We evaluate our method on well-known large-scale metagenomic studies and show that our proposed approach outperforms comparative state-of-the-art methods for disease prediction.

Speakers:

Mohammad Rahman

George Mason University

05:00 PM - 05:20 PM (EDT)

MLCSB: Machine Learning - Towards Heterogeneous Information Fusion: Bipartite Graph Convolutional Networks for In Silico Drug Repurposing

Motivation: Mining disease and drug association and their interactions are essential for developing computational models in drug repurposing and understanding underlying biological mechanisms. Recently, large-scale biological databases are increasingly available for pharmaceutical research, allowing for deep characterization for molecular informatics and drug discovery. In this study, we propose a bipartite graph convolution network model that integrates multi-scale pharmaceutical information. Especially the introduction of protein nodes serve as a bridge of message passing, which provides insights into the protein-protein interaction (PPI) network for improved drug repositioning assessment. Results: Our approach combines insights of multi-scale pharmaceutical information by constructing a multi-relational graph of protein–protein, drug–protein and disease-protein interactions. Specifically, our model offers a novel avenue for message passing among diverse domains that we learn useful feature representations for all graph nodes by fusing biological information across interaction edges. Then the high-level representation of drug-disease pairs are fed into a multi-layer perceptron decoder to predict therapeutic indications. Unlike conventional graph convolution networks that assume the same node attributes in a global graph, our model is domain-consistent by modeling inter-domain information fusion with bipartite graph convolution operation. We offered an exploratory analysis for finding novel drug-disease associations. Extensive experiments showed that our approach achieves improved performance than multiple baseline approaches.

Speakers:

Zichen Wang

UCLA

05:00 PM - 05:20 PM (EDT)

SST01 - Interpreting genetic variants through chromatin interaction maps in primary human immune cells

Speakers:

Ferhat Ay

La Jolla Institute for Immunology & UCSD

05:00 PM - 05:55 PM (EDT)

FILLING THE GAPS IN ONLINE LEARNING

Group Think Tank - discussion: What is Missing from Online Tools in Support of Online Learning

Speakers:

Patricia Palagi

SIB Swiss Institute of Bioinformatics

05:00 PM - 05:10 PM (EDT)

SST02 - Citizen Science and Videogames - An unlikely marriage

We dedicated the last five years at Massively Multiplayer Online Science to bring citizen science activities to major videogames that resulted in the most active citizen science projects. We'll share the major milestones of this journey from Project Discovery Proteins to Borderlands Science and on.

Speakers:

Attila Szantner

05:10 PM - 05:20 PM (EDT)

SST02 - Challenges in designing scientific discovery games in AAA games

How do we make a complex scientific task suitable to the universe of a sci-fi fast-paced role-playing first person shooter video game? In this talk, we will describe the challenges and choices that were made when designing the Borderlands Science puzzle game. We will also discuss how scientists and video game designers can efficiently work together to make a citizen science game scientifically relevant and fun!

Speakers:

Gabriel Richard

Gearbox Studio Quebec

05:20 PM - 05:40 PM (EDT)

CAMDA - DNA Based Methods in Intelligence - Moving Towards Metagenomics

TBP

Speakers:

Paweł Łabaj

Małopolska Centre of Biotechnology of Jagiellonian University

05:20 PM - 05:30 PM (EDT)

SST02: TBA

Speakers:

Haukur Pálsson

CCP Games

05:20 PM - 05:50 PM (EDT)

SST01 - Exploring the functional consequences of macrophage heterogeneity

Speakers:

Kathryn Miller-Jensen

Yale University

05:20 PM - 05:40 PM (EDT)

iRNA - Finding the Direct Optimal RNA Barrier Energy and Improving Pathways with an Arbitrary Energy Model

RNA folding kinetics plays an important role in the biological functions of RNA molecules. An important goal in the investigation of the kinetic behavior of RNAs is to find the folding pathway with the lowest energy barrier. For this purpose, most of the existing methods employ heuristics because the number of possible pathways is huge even if only the shortest (direct) folding pathways are considered. In this study, we propose a new method using a best-first search strategy to efficiently compute the exact solution of the minimum barrier energy of direct pathways. Using our method, we can find the exact direct pathways within a Hamming distance of 20, while the previous methods even miss the exact short pathways. Moreover, our method can be used to improve the pathways found by existing methods for exploring indirect pathways. The source code and datasets created and used in this research are available at https://github.com/eukaryo/czno.

Speakers:

Hiroki Takizawa

The University of Tokyo

05:20 PM - 05:40 PM (EDT)

SysMod - Modeling Sorting, Intercalation, and Involution Tissue Behaviors due to Regulated Cell Adhesion

Cell-cell adhesion can dictate tissue growth and multicellular pattern formation and it is crucial for the cellular dynamics during embryogenesis and cancer progression. While it is known that these adhesive forces are generated by cell adhesion molecules (CAMs), the regulation of CAMs is not well understood due to complex nonlinear interactions that span multiple levels of biological organization–from genetic regulation to whole-organism shape formation. We present a novel continuous model that can explain the dynamic relationships between genetic regulation, CAM expression, and differential adhesion. This approach can demonstrate the mechanisms responsible for cell-sorting behaviors, cell intercalation in proliferating populations, and the involution of zebrafish germ layer cells during gastrulation. The model can predict the physical parameters controlling the amplitude and wavelength of a cellular intercalation interface as shown in vitro. We demonstrate the crucial role of N-cadherin regulation for the involution and migration of cells beyond the gradient of the morphogen Nodal during zebrafish gastrulation. Integrating the emergent spatial tissue behaviors with the regulation of genes responsible for essential cellular properties such as adhesion will pave the way toward understanding the genetic regulation of large-scale complex patterns and shapes formation in developmental, regenerative, and cancer biology.

Speakers:

Jason Ko

UMBC

05:20 PM - 05:40 PM (EDT)

Text Mining - Deep Semi-supervised Ensemble Method for Classifying Co-mentions of Human Proteins and Phenotypes

Identifying human protein-phenotype relations is of paramount importance for uncovering rare and complex diseases. Human Phenotype Ontology (HPO) is a standardized vocabulary for describing disease-related phenotypic abnormalities in humans. Since the experimental determination of HPO categories for human proteins is a highly resource-consuming task, developing automated tools that can accurately predict HPO categories has gained interest recently. In this work, we develop a novel method for classifying sentence co-mentions of human proteins and HPO names using semi-supervised learning and deep neural networks. Our model is a combination of BERT, CNN, RNN models, and a self-learning module, that uses a large collection of unlabeled co-mentions available from ProPheno, which is an online database developed in a previous study. Using a gold-standard dataset composed of curated sentence-level co-mentions, we demonstrate that our proposed model provides the state-of-the-art performance in the task of classifying human protein-phenotype co-mentions by outperforming two supervised and semi-supervised support vector machines model counterparts. The findings and the insight of this work have implications for biocurators, researchers, and bio text mining tool developers.

Speakers:

Morteza Shahri

Montana State University

05:20 PM - 05:30 PM (EDT)

MLCSB: Machine Learning - Learning Context-aware Structural Representations to Predict Antigen and Antibody Binding Interfaces

Understanding how antibodies specifically interact with their antigens can enable better drug and vaccine design, as well as provide insights into natural immunity. Experimental structural characterization can detail the “ground truth” of antibody-antigen interactions, but computational methods are required to efficiently scale to large-scale studies. In order to increase prediction accuracy as well as to provide a means to gain new biological insights into these interactions, we have developed PECAN, a unified deep learning-based framework to predict binding interfaces on both antibodies and antigens. PECAN leverages three key aspects of antibody-antigen interactions to learn predictive structural representations: (1) since interfaces are formed from multiple residues in spatial proximity, we employ graph convolutions to aggregate properties across local regions in a protein; (2) since interactions are specific between antibody-antigen pairs, we employ an attention layer to explicitly encode the context of the partner; (3) since more data is available for general protein-protein interactions, we employ transfer learning to leverage this data as a prior for the specific case of antibody-antigen interactions. We show that PECAN achieves state-of-the-art performance at predicting binding interfaces on both antibodies and antigens, and that each of its three aspects drives additional improvement in the performance.

Speakers:

Srivamshi Pittala

Dartmouth College

05:20 PM - 05:40 PM (EDT)

Function - eCAMI: simultaneous classification and motif identification for enzyme annotation

Carbohydrate-active enzymes (CAZymes) are extremely important to bioenergy, human gut microbiome, and plant pathogen researches and industries. We developed a new amino acid k-mer based CAZyme classification, motif identification, and genome annotation tool using a bipartite network algorithm. Using this tool, we classified 390 CAZyme families into thousands of subfamilies each with distinguishing k-mer peptides. These k-mers represented the characteristic motifs of each subfamily, and thus were further used to annotate new genomes for CAZymes. This idea was also generalized to extract characteristic k-mer peptides for all the Swiss-Prot enzymes classified by the EC (enzyme commission) numbers and applied to enzyme EC prediction. This new tool was implemented as a Python package named eCAMI. Benchmark analysis of eCAMI against the state-of-the-art tools on CAZyme and enzyme EC datasets found that: (i) eCAMI has the best performance in terms of accuracy and memory use for CAZyme and enzyme EC classification and annotation; (ii) the k-mer based tools (including PPR-Hotpep, CUPP, eCAMI) perform better than homology-based tools and deep-learning tools in enzyme EC prediction. Lastly, we confirmed that the k-mer based tools have the unique ability to identify the characteristic k-mer peptides in the predicted enzymes. The paper is published at https://doi.org/10.1093/bioinformatics/btz908.

Speakers:

Yanbin Yin

Department of Food Science and Technology, Nebraska Food for Health Center, University of Nebraska-Lincoln

05:30 PM - 06:00 PM (EDT)

SST02: Panel Discussion

TBD

Speakers:

All Speakers

05:30 PM - 05:40 PM (EDT)

MLCSB: Machine Learning - PaccMann^RL: Designing anticancer drugs from transcriptomic data via reinforcement learning

While state-of-the-art deep learning approaches have shown potential in generating compounds with desired chemical properties, they disregard the cellular biomolecular properties of the target disease. We introduce a novel framework for de-novo molecular design that systematically leverages systems biology information into the drug discovery process. Embodied through two separate VAEs, the drug generation is driven by a disease context (transcriptomic proﬁles of cancer cells) deemed to represent the target environment of the drug. Showcased at the task of anticancer drug discovery, our conditional generative model is demonstrated to tailor anticancer compounds to target desired biomolecular proﬁles. Specifically, we reveal how the molecule generation can be biased towards compounds with high predicted inhibitory eﬀect against individual cell-lines or cell-lines from speciﬁc cancer sites. We verify our approach by investigating candidate drugs generated against speciﬁc cancer types and ﬁnd highest structural similarity to existing compounds with known eﬃcacy against these types. Despite no direct optimization of other pharmacological properties, we report good agreement with cancer drugs in metrics like drug-likeness, synthesizability and solubility. We envision our approach to be a step towards increasing success rates in lead compound discovery and ﬁnding more targeted medicines by leveraging the cellular environment of the disease.

Speakers:

Jannis Born

ETH Zurich

05:40 PM - 06:00 PM (EDT)

iRNA - The locality dilemma of Sankoff-like RNA alignments

Motivation: Elucidating the functions of non-coding RNAs by homology has been strongly limited due to fundamental computational and modeling issues. While existing simultaneous alignment and folding (SA&F) algorithms successfully align homologous RNAs with precisely known boundaries (global SA&F), the more pressing problem of finding homologous RNAs in the genome (local SA&F) is intrinsically more difficult and much less understood. Typically, the length of local alignments is strongly overestimated and alignment boundaries are dramatically mispredicted. We hypothesize that local SA&F approaches are compromised this way due to a score bias, which is caused by the contribution of RNA structure similarity to their overall alignment score. Results: In the light of this hypothesis, we study local SA&F for the first time systematically—based on a novel local RNA alignment benchmark set and quality measure. First, we vary the relative influence of structure similarity compared to sequence similarity. Putting more emphasis on the structure component leads to overestimating the length of local alignments. This clearly shows the bias of current scores and strongly hints at the structure component as its origin. Second, we study the interplay of several important scoring parameters by learning parameters for local and global SA&F. The divergence of these optimized parameter sets underlines the fundamental obstacles for local SA&F. Thirdly, by introducing a position-wise correction term in local SA&F, we constructively solve its principal issues.

Speakers:

Teresa Müller

University Freiburg, Bioinformatics

05:40 PM - 06:00 PM (EDT)

SysMod - Closing remarks of the first SysMod day

A brief recap of the first day of the meeting.

Speakers:

Juilee Thakar

University of Rochester

05:40 PM - 06:00 PM (EDT)

CAMDA - Day summary

Speakers:

Joaquin Dopazo

Clinical Bioinformatics Area. Fundación Progreso y Salud, Sevilla

05:40 PM - 06:00 PM (EDT)

Text Mining - A Hybrid Method for Phenotype Concept Recognition using the Human Phenotype Ontology

Automatic phenotype concept recognition from unstructured text remains a challenging task in biomedical text mining research despite a few attempts in the past. In previous works, dictionary-based and machine learning-based methods are attempted for phenotype concept recognition. The dictionary-based methods can achieve a high precision, but it suffers from a lower recall problem. Machine learning-based methods can recognize more phenotype concept variants by automatic feature learning. However, most of them require large corpora of manually annotated data for its model training, which is difficult to obtain due to the high cost of human annotation. To address the problems, we propose a hybrid method combining dictionary-based and machine learning-based methods to recognize the Human Phenotype Ontology (HPO) concepts in unstructured biomedical text. We firstly use all concepts and synonyms in HPO to construct a dictionary. Then we employ the dictionary-based method and HPO to automatically build a “weak” training dataset for machine learning. Next, a cutting-edge deep learning model is trained to classify each candidate phrase into a corresponding concept id. Finally, machine learning-based prediction results are incorporated into the dictionary-based results for improved performance. Experimental results show that our method can achieve state-of-the-art performance without any manually labeled training data.

Speakers:

Ling Luo

National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health

05:40 PM - 05:50 PM (EDT)

MLCSB: Machine Learning - Tissue-guided LASSO for prediction of clinical drug response using preclinical samples

Predicting the clinical drug response (CDR) of cancer patients, based on their clinical parameters and their tumours' molecular profiles, can play an important role in precision medicine. While machine learning (ML) models have the potential to address this issue, their training requires data from a large number of patients treated with each drug, limiting their feasibility for many drugs. One alternative is training ML models on large databases containing molecular profiles of hundreds of preclinical cell lines and their response to hundreds of drugs. Here, we developed a novel algorithm (TG-LASSO) that explicitly incorporates information on samples' tissue of origin with gene expression profiles to predict the CDR of patients using preclinical samples. Using two large databases, we showed that TG-LASSO can accurately distinguish between resistant and sensitive patients for 7 out of 12 drugs, outperforming various other methods. Moreover, TG-LASSO identified genes associated with the drug response, including known targets and pathways involved in the drugs' mechanism of action. Additionally, genes identified by this method for multiple drugs in a tissue are associated with patient survival and can be used to predict their outcome. In summary, TG-LASSO can predict patients’ CDR and identify biomarkers of drug sensitivity and survival.

Speakers:

Amin Emad

McGill University

05:40 PM - 06:00 PM (EDT)

Function - Learning sequence, structure and network features for protein function prediction

Recently, self-supervised and unsupervised representation learning approaches have shown tremendous potential in exploring the huge volume of protein data in many databases and learning features indicative of protein function. Building upon our previous works on sequence, structure, and network data, we propose to systematically investigate the contribution of these features on classification performance on individual GO terms, study their complementarity, and build an integrative model. We compute: sequence-based features using our LSTM language model pre-trained on ~10 million protein sequences; structure-based features from contact maps using a Graph Autoencoder pretrained on ~30k domain structures from CATH and network-based features using our deepNF model pre-trained on 6 different networks from STRING. Then we train a separate neural network (NN) for predicting GO term probabilities on each individual feature set. We will show results from an experiment we conducted on ~18,000 human proteins using their sequences, 3D structures retrieved from PDB and SWISS-MODEL and PPI networks. We will also present the performance results obtained by training a multi-modal NN with all three feature sets for multiple organisms. We show that different modalities contribute to different GO terms, and show that the model integrating information from all sources outperforms the individual models.

Speakers:

Meet Barot

Center for Data Science, New York University

05:50 PM - 06:00 PM (EDT)

SST01 - Dissecting the heterogeneity of protein and transcriptional responses in human blood derived immune cells after T- and monocyte-specific activation

Single cell profiling of activated immune cells is a powerful way to study immune function disruptions in diverse diseases and conditions. However, generating and analyzing single cell data from activated cells harbor unique challenges. To address these challenges and derive reference response genes to frequently studied immune cell activation conditions, we generated CITE-seq data from 10 healthy adults (5 men, 5 women) before and after stimulating their peripheral blood mononuclear cells (PBMCs) via i) anti-CD3/CD28 and ii) LPS. Using a comprehensive antibody panel (n=39) including cell type (e.g., CD16, CD14) and cell state (e.g., CD69, CD25) markers, we uncovered how each immune cell type responds to these conditions. Two levels of multiplexing (cell hashing and individuals’ genotypes) were instrumental to avoid batch effects and to eliminate multiplet cells that increased in number upon activation. All PBMCs responded to stimulation via anti-CD3/CD28, whereas LPS specifically activated monocytes and induced pro-inflammatory genes and pathways. Pseudo-temporal ordering of single cells before and after activation revealed cell- and condition-specific heterogeneity in cellular responses independent of genetic variation. Together, these data are shared within an interactive web application (https://czi-pbmc-cite-seq.jax.org/) and will serve as a resource to guide future studies of immune cell responses.

Speakers:

Nathan Lawlor

The Jackson Laboratory

05:50 PM - 06:00 PM (EDT)

MLCSB: Machine Learning - Quantifying gene selection in cancer through protein functional alteration bias

Compiling the catalogue of genes actively involved in cancer is an ongoing endeavor, with profound implications to the understanding and treatment of the disease. An abundance of computational methods have been developed to screening the genome for candidate driver genes based on genomic data of somatic mutations in tumors. Existing methods make many implicit and explicit assumptions about the distribution of random mutations. We present FABRIC, a new framework for quantifying the selection of genes in cancer by assessing the effects of de-novo somatic mutations on protein-coding genes. Using a machine-learning model, we quantified the functional effects of ∼3M somatic mutations extracted from over 10 000 human cancerous samples, and compared them against the effects of all possible single-nucleotide mutations in the coding human genome. We detected 593 protein-coding genes showing statistically significant bias towards harmful mutations. These genes, discovered without any prior knowledge, show an overwhelming overlap with known cancer genes, but also include many overlooked genes. FABRIC is designed to avoid false discoveries by comparing each gene to its own background model using rigorous statistics, making minimal assumptions about the distribution of random somatic mutations. The framework is an open-source project with a simple command-line interface.

Speakers:

Nadav Brandes

The Hebrew University of Jerusalem

05:55 PM - 06:00 PM (EDT)

Web - Closing of WEB 2020

Speakers:

Michelle Brazas

Ontario Institute for Cancer Research

06:00 PM - 06:10 PM (EDT)

SST01 - Single-cell transcriptomic analysis of SARS-CoV-2 reactive CD4+ T cells

The contribution of CD4+ T cells to protective or pathogenic immune responses to SARS-CoV-2 infection remains unknown. Here, we present large-scale single-cell transcriptomic analysis of viral antigen-reactive CD4+ T cells from 32 COVID-19 patients. In patients with severe disease compared to mild disease, we found increased proportions of cytotoxic follicular helper (TFH) cells and cytotoxic T helper cells (CD4-CTLs) responding to SARS-CoV-2, and reduced proportion of SARS-CoV-2 reactive regulatory T cells. Importantly, the CD4-CTLs were highly enriched for the expression of transcripts encoding chemokines that are involved in the recruitment of myeloid cells and dendritic cells to the sites of viral infection. Polyfunctional T helper (TH)1 cells and TH17 cell subsets were underrepresented in the repertoire of SARS-CoV-2-reactive CD4+ T cells compared to influenza-reactive CD4+ T cells. Together, our analyses provide so far unprecedented insights into the gene expression patterns of SARS-CoV-2 reactive CD4+ T cells in distinct disease severities.

Speakers:

Benjamin Meckiff

La Jolla Institute for Immunology

06:00 PM - 07:00 PM (EDT)

iRNA - iRNA: Social Hour

Social Hour

Speakers:

Klemens Hertel

UC Irvine

Michelle Scott

Université de Sherbrooke

Yoseph Barash

University of Pennsylvania

06:10 PM - 06:20 PM (EDT)

SST01 - Single-cell transcriptomic analysis of allergen-specific T cells in allergy and asthma

CD4+ helper and regulatory T cells that respond to common allergens play an important role in driving and dampening airway inflammation in patients with asthma. Until recently, direct, unbiased molecular analysis of allergen-reactive T-cells has not been possible. To better understand the diversity of these T-cells in allergy and asthma, we analyzed the single-cell transcriptome of ~50,000 house dust mite (HDM) allergen-reactive T cells from asthmatics with HDM allergy and from three control groups: asthmatics without HDM allergy and non-asthmatics with and without HDM allergy. Our analyses show that HDM allergen-reactive T cells are highly heterogeneous, and certain subsets are quantitatively and qualitatively different in subjects with HDM-reactive asthma. The number of interleukin-9 expressing HDM-reactive TH cells is greater in asthmatics compared with non-asthmatics with HDM allergy and display enhanced pathogenic properties. More HDM-reactive TH and Treg cells expressing the interferon-response signature are present in asthmatics without HDM allergy. In cells from these subsets, expression of TNFSF10 was enriched; its product, TRAIL, dampens activation of TH cells. These findings suggest that these subsets may dampen allergic responses, which may help explain why only some people develop TH2 responses to nearly ubiquitous allergens.

Speakers:

Gregory Seumois

La Jolla Institute for Immunology

09:30 AM - 10:30 AM (EDT)

Keynotes - Computational modeling of tumor immunity

Speakers:

Xiaole Liu

Dana-Farber Cancer Institute / Harvard University

10:40 AM - 11:20 AM (EDT)

Function - Keynote: Gene function prediction using unsupervised biological network integration

Biological networks have the power to map cellular function, but only when unified to overcome their individual limitations such as bias and noise. Unsupervised network integration addresses this, automatically weighting input information to obtain an accurate, unified result. However, existing unsupervised network integration methods do not adequately scale to the number of nodes and networks present in genome-scale data and do not handle frequently encountered data characteristics (e.g. partial network overlap). To address this, we have developed an unsupervised deep learning-based network integration algorithm that incorporates recent advances in reasoning over unstructured data – namely the Graph Convolutional Network (GCN) – that can effectively learn dependencies between physical, co-expression and genetic interaction network topologies. Our method, BIONIC (Biological Network Integration using Convolutions), produces high quality gene and protein features which capture and unify information across many diverse functional interaction networks. BIONIC learns features which contain substantially more functional information compared to existing approaches, linking genes and proteins that share co-complex, pathway and bioprocess relationships.

Speakers:

Gary Bader

University of Toronto

10:40 AM - 11:00 AM (EDT)

iRNA - Detection of differential RNA modifications from direct RNA sequencing of human cell lines

Differences in RNA expression can provide insights into the molecular identity of a cell and pathways involved in human diseases. RNA modifications such as m6A have been found to contribute to molecular functions of RNAs. However, quantification of differences in RNA modifications has been challenging. Here, we present a computational method, xPore, to identify differential RNA modifications from direct RNA sequencing data. Based solely on the current intensity profiles, we extend a standard two-Gaussian mixture model to accommodate multi-sample comparisons. For each single site, the model learns two distributions, corresponding to unmodified and modified RNA, the signal properties that are shared across samples, while allowing the probability of each read being modified to be inferred specifically for each sample. Having incorporated prior knowledge into the model, we are able to determine the signal distributions of the modified k-mers and quantitatively estimate the modification rates accordingly. We evaluate our method on transcriptome-wide m6A profiling, demonstrating that we can accurately prioritize differentially modified sites. Together, we demonstrate that RNA modifications can be quantitatively identified from direct RNA-sequencing data with high accuracy, opening many new opportunities for large scale applications in precision medicine. xPore is available at https://github.com/GoekeLab/xpore.

Speakers:

Ploy Pratanwanich

Genome Institute of Singapore

10:40 AM - 11:00 AM (EDT)

SysMod - A stochastic hybrid model for DNA replication incorporating protein mobility dynamics

DNA replication is a complex process that ensures genetic information maintenance. As recently observed, DNA replication timing is highly correlated with chromatin folding and global nuclear architecture. Here, we present a stochastic hybrid model of DNA replication that incorporates protein mobility and three-dimensional chromatin structure. Our model provides a framework to realistically simulate DNA replication for a complete eukaryotic genome and investigate the relationship between three-dimensional chromatin conformation and replication timing. Performing simulations for three model variants and a broad range of parameter values, we collected about 300,000 in silico replication profiles for fission yeast. We find that the number of firing factors initiating replication is rate-limiting and dominates the DNA replication completion time. We also find that explicitly modeling the recruitment of firing factors by the spindle pole body best reproduces experimental data and provide an independent validation of these findings in vivo. Further investigation of replication kinetics confirmed earlier observations of a rate-limiting number of firing factors in conjunction with their recycling upon replication fork collision. While the model faithfully reproduces global patterns of replication initiation, additional analysis of firing concurrence suggests that a uniform binding probability is too simplistic to capture local neighborhood effects in origin firing.

Speakers:

Maria Anna Rapsomaniki

IBM Research Zurich

10:40 AM - 11:00 AM (EDT)

Evolution and Comparative Genomics - FastMulRFS: Fast and accurate species tree estimation under generic gene duplication and loss models

Motivation:Species tree estimation is a basic part of biological research but can be challenging because of gene duplication and loss (GDL), which results in genes that can appear more than once in a given genome. All common approaches in phylogenomic studies either reduce available data or are error-prone, and thus, scalable methods that do not discard data and have high accuracy on large heterogeneous datasets are needed. Results: We present FastMulRFS, a polynomial-time method for estimating species trees without knowledge of orthology. We prove that FastMulRFS is statistically consistent under a generic model of GDL when adversarial GDL does not occur. Our extensive simulation study shows that FastMulRFS matches the accuracy of MulRF (which tries to solve the same optimization problem) and has better accuracy than prior methods, including ASTRAL-multi (the only method to date that has been proven statistically consistent under GDL), while being much faster than both methods.

Speakers:

Erin Molloy

University of California, Los Angles

10:40 AM - 11:00 AM (EDT)

SST03 - Tackling the challenge of elucidating reef-building coral holobiont function

Reef-building corals form massive calcium carbonate structures that are visible from space and support invaluable reef goods and services. These corals are metaorganisms whose functionality comes from the holobiont integration of the cnidarian host, its primary dinoflagellate endosymbionts (family Symbiodinaceae), and a multitude of bacteria, archaea, viruses, and fungi that are surface, tissue, and skeletal dwelling symbionts. The study of this symbiotic system as a fascinating functional puzzle has intensified with the rapid decline in coral cover and growing threat of reef extinction due to climate change factors, such as increasing seawater temperatures and ocean acidification. In particular, the high frequency of mass bleaching events, where thermal stress causes dysbiosis and loss or expulsion the Symbiodinaceae and mass coral mortality, has sharpened our focus on these threatened organisms. Technological advances have allowed a rapid expansion of coral and symbiont sequence data for studies of coral response to stress, but bioinformatic challenges remain in analyzing and interpreting this wealth of metaorganism data.

Speakers:

Hollie Putnam

University of Rhode Island

10:40 AM - 11:00 AM (EDT)

ODSS - Introduction to the ODSS Data Science Sessions

Speakers:

Patti Brennan

National Library of Medicine, NIH

Susan Gregurick

Associate Director for Data Science and Director of the NIH Office for Data Science Strategy

10:40 AM - 11:20 AM (EDT)

Education - COSI Education Keynote Talk: Empowering usable, and comprehensive bioinformatics training

With the explosion of biological data, the primary challenge is not how to store the data, nor what computational resources to use to process them, but it is the general lack of researchers' understanding to manipulate and analyse these data. This problem could be solved with comprehensive and up-to-date bioinformatics training. How can we build such an infrastructure? Over the last few years, I had the opportunity to lead and contribute to several training programs and communities in bioinformatics. In this talk, I will share some lessons learned and ideas to build usable, comprehensive, and empowering bioinformatics training infrastructure. I will talk about community-driven development of free, FAIR and reusable material, online and hybrid training, mentoring, and further illustrate these concepts with concrete examples from my work.

Speakers:

Bérénice Batut

University of Freiburg

10:40 AM - 11:40 AM (EDT)

MLCSB: Machine Learning - Finding concise descriptors of genomic data

Genome scale technologies can measure thousands or even millions of molecules but the individual measurements are often noisy redouts of the internal biological state. For many computational approaches the ultimate goal is to infer the true biological state parameters from their effects on a much larger set of experimental readouts. Indeed, much of genomic data analysis can be viewed as a special case of dimensionality reduction. Unlike a generic dimensionality reduction problem, however, in the biological case we have extensive prior knowledge about the data-generating process. In this talk we will discuss several methods that exploit this prior knowledge to create interpretable low-dimensional representations for genome scale datasets.

Speakers:

Maria Chikina

University of Pittsburgh

10:45 AM - 11:40 AM (EDT)

CAMDA - CAMDA KEYNOTE: Challenges and tools for integrative analysis of single cell 'omics data

Single-cell multi 'omics data has potential to provide unprecedented insights into molecular, spatial and cellular organization of tissue during health and disease. Whilst approaches for integrating single-cell data 'omics are emerging, the sparsity, high dimensionality, heterogeneity and scale of these data type present unique challenges. Due to the scale of single cell data, dimension reduction is an essential preliminary step in analysis pipelines, data integration, cell clustering and trajectory analysis. Principal component analysis (PCA) is a widely used for this because it is relatively fast, and can easily scale to large datasets when used with sparse-matrix representations. I will review different forms of PCA, impact of scaling, log-transforming, and how to recognize problems such as horseshoe or arch effect. I will describe a alternative to PCA which we have recently implemented in a new Bioconductor package Corral, and show how simple replacement of PCA with corral can improve integrative analysis of single cell data. I will describe challenges we have encounters and how we are re-envisioning tools we developed for bulk RNAseq data for emerging single cell data types.

Speakers:

Aedin Culhane

DANA-FARBER CANCER INST

11:00 AM - 11:20 AM (EDT)

iRNA - RNA editing landscapes: a new model for biomarkers discovery in neurological disease

Current studies look at genomic variants or differential gene expression to predict genetic predisposition and find biomarkers for many neurodevelopmental, psychiatric and degenerative disorders. Here we explore a novel approach to elucidate potential diagnostic, prognostic and/or therapeutic biomarkers for major depressive disorder and suicide focusing on a more nuanced aspect of transcriptome diversity, RNA editing. RNA editing, more specifically adenosine deaminase acting on RNA (ADAR) editing which contributes to transcriptome diversity by dynamically altering the ratios of differentially functioning proteins unpinning the “fine-tuning” of neural signaling and synaptic plasticity. We apply a novel approach utilizing an item response theory model the Guttman Scale to create ADAR editing landscape profiles which are then used to map differential editing in major depressive disorder and suicide. We were able to find a handful of genes of interest for further investigation into their direct contribution to neurological symptoms. We also highlight pathways including ion homeostasis that are altered in depression and suicide which warrant further investigation for their role in synaptic plasticity. Furthermore, we provide evidence this model can be used in biomarker discovery for many other neurological disorders in which transcriptome diversity plays an important role.

Speakers:

Noel-Marie Plonski

Kent State University

11:00 AM - 11:40 AM (EDT)

SysMod - Quantitative and Systems Pharmacology (QSP) and Model-Informed Drug Development (MIDD) of a “Smart” Insulin

One of the many challenges of drug development programs is the strictly limited volume of pharmacological data for investigatory drug candidates. To help manage the risk of exposing volunteers to drug candidates of unknown safety and efficacy while evaluating hypothesized benefits, programs seek to integrate all relevant data including molecular and cellular profiling data and clinical trial electronic data capture. The integration of relevant data has been used successfully to help guide decisions at each step of the development process. The success of such efforts across the industry is evidenced by the FDA’s launch of a MIDD pilot program to facilitate the use of quantitative methods to ultimately help improve clinical trial efficiency and optimize therapeutic individualization in the absence of dedicated trials. The integration of quantitative methods was used to guide validation of putative glucose dysregulation targets and to guide the design of a complex clinical trial of glucose control. Although every program has individual challenges, these case studies will be discussed to highlight open questions in the application of quantitative systems biology in drug development.

Speakers:

Carolyn Cho

11:00 AM - 11:20 AM (EDT)

SST03 - Dynamic genome evolution of coral symbionts: a bioinformatics challenge

Coral reefs are hotspots of marine biodiversity. The ecological success of corals in nutrient-poor waters relies on photosynthetic algal symbionts (Symbiodiniaceae) for supply of fixed carbon as energy, and nutrients. However, the implication of the evolution of these algae and its implications on coral evolution is poorly understood. Genomes of Symbiodiniaceae present a bioinformatics challenge, because of their immense sizes (1-3 Gbp) and highly idiosyncratic genome features. In this talk, I will present our recent effort to generate 12 de novo genome assemblies from diverse Symbiodiniaceae species and their free-living relative, and how we developed a customised computational workflow for predicting genes from these genomes. Comparative analysis reveals unexpectedly high sequence and structural divergence, and conserved lineage-specific gene families of unknown function. Many genes are encoded in unidirectional clusters, and some critical functions are encoded in tandemly repeated single-exon genes. Our results underscore the rapid evolution of coral symbionts that comprise a massive, undiscovered phylogenetic diversity, and elucidate how selection acts within the context of a complex genome structure to facilitate local adaptation, e.g. by enhancing transcriptional responses. These outcomes provide an important reference for research of coral holobionts and their resilience in changing environments.

Speakers:

Cheong Chan

University of Queensland

11:00 AM - 11:20 AM (EDT)

10 lessons we learned in 10 days about clinical data interoperability in the COVID crisis

There is an urgent need to take what we have learned in our new data-driven era of medicine, and use it to create a new system of precision medicine, delivering the best, safest, cost-effective preventative or therapeutic intervention at the right time, for the right patients. Dr. Butte's lab at the University of California, San Francisco builds and applies tools that convert trillions of points of molecular, clinical, and epidemiological data -- measured by researchers and clinicians over the past decade and now commonly termed “big data” -- into diagnostics, therapeutics, and new insights into disease. Dr. Butte, a computer scientist and pediatrician, will highlight his center’s recent work on integrating electronic health records data across the entire University of California, and how analytics on this “real world data” can lead to new evidence for drug efficacy, new savings from better medication choices, and new methods to teach intelligence – real and artificial – to more precisely practice medicine.

Speakers:

Atul Butte

UCSF

11:00 AM - 11:10 AM (EDT)

Evolution and Comparative Genomics - Species tree estimation with selection

Species trees form a basis to address many biological questions, and during the last years have crucial to our understanding of the divergence process and speciation. Although modeling the species tree is challenging the last two decades have seen an explosion of sophisticated models for the species tree. The multispecies coalescent model has arisen as a leading framework, but alternative ones have been proposed. The polymorphism-aware phylogenetic models (PoMos), in particular, offer the advantage of naturally accounting for incomplete lineage sorting and being flexible modeling-wise. PoMo extends any DNA substitution model to account for polymorphisms by expanding the state space to include polymorphic states. A Moran process is used to model genetic drift. PoMo performs well for species tree estimation as it can accurately and time-efficiently estimate the parameters describing evolutionary patterns for phylogenetic trees of any shape (species trees, population trees, or any combination of those). Here, we extend PoMo to incorporate allelic directional selection. Our results show that one might overestimate the divergence among evolving species when directional selection is unaccounted. We developed a Bayesian framework for PoMo with allelic selection, which permitted to infer pervasive signatures of nucleotide usage biases in great apes and fruit flies genomes.

Speakers:

Carolin Kosiol

University of St Andrews

11:10 AM - 11:20 AM (EDT)

Evolution and Comparative Genomics - Graph Splitting: A Graph-Based Approach for Superfamily-Scale Phylogenetic Tree Reconstruction

A protein superfamily contains distantly related proteins that have acquired diverse biological functions through a long evolutionary history. Phylogenetic analysis of the early evolution of protein superfamilies is a key challenge because existing phylogenetic methods show poor performance when protein sequences are too diverged to construct an informative multiple sequence alignment (MSA). Here, we propose the Graph Splitting (GS) method, which rapidly reconstructs a protein superfamily-scale phylogenetic tree using a graph-based approach. Evolutionary simulation showed that the GS method can accurately reconstruct phylogenetic trees and be robust to major problems in phylogenetic estimation, such as biased taxon sampling, heterogeneous evolutionary rates, and long-branch attraction when sequences are substantially diverge. Its application to an empirical data set of the triosephosphate isomerase (TIM)-barrel superfamily suggests rapid evolution of protein-mediated pyrimidine biosynthesis, likely taking place after the RNA world. Furthermore, the GS method can also substantially improve performance of widely used MSA methods by providing accurate guide trees. The GS method is freely available at our website: http://gs.bs.s.u-tokyo.ac.jp/ Reference: Motomu Matsui and Wataru Iwasaki. Systematic Biology, 69, 265-279. (2020) https://doi.org/10.1093/sysbio/syz049

Speakers:

Wataru Iwasaki

The University of Tokyo

11:20 AM - 11:40 AM (EDT)

Function - Cross-species functional prediction by global network alignment

We report the first successful cross-species pre- diction of protein function based solely on topology-driven global network alignment. Using SANA (the Simulated Annealing Network Aligner), we pair the proteins of one species with those of another solely by maximizing the number of aligned edges between the networks. We find that SANA’s confidence, called NAF, in each individual pair of aligned proteins correlates with their functional similarity. We then apply SANA to BioGRID 3.0 networks from April 2010, and use GO data from the same month to transfer GO annotations from better-annotated proteins to lesser-annotated ones. We validate the predictions on a recent GO release and find an AUPR of up to 0.4 depending on the predicting GO evidence code, even when restricting predictions to proteins that have no observed sequence or homology relationship. Finally, we apply the same method to recent BioGRID PPI networks of mouse and human, and predict novel cilia-related GO terms in human proteins based on their confident alignment with cilia-annotated mouse proteins; the most confident predictions have literature validation rates above 80%. We propose topology-based alignment of PPI networks as a novel source for prediction of protein function that is independent of sequence or structural information

Speakers:

Wayne Hayes

UCI

University of Montreal

11:30 AM - 02:00 PM (EDT)

iRNA - iRNA: Poster Session

Poster session going through lunch

Speakers:

Klemens Hertel

UC Irvine

Michelle Scott

Université de Sherbrooke

Yoseph Barash

University of Pennsylvania

12:00 PM - 12:20 PM (EDT)

SST03 - Metabolic complementary in phylogenetically divergent coral holobionts

Coral holobiont members – which include the coral host, the algal symbiont and a poorly characterized consortium of diverse microbial taxa – have different generation times and effective population sizes. These differences can have pronounced evolutionary and ecological consequences for holobiont members, especially under a rapidly changing climate. Corals tend to have slow generation times relative to their microbes, on the order of decades. Algal symbionts with different physiological adaptations to temperature are known to shift in abundance within a host in response to perturbations (e.g., global warming) over months to years. Coral associated prokaryotic microbial assemblages not only have very fast generation times, but their community composition can shift rapidly under changing environmental conditions. Metabolic complementary analysis has revealed a keystone role for the coral microbiome in holobiont homeostasis. We performed temperature stress experiments on three Caribbean coral holobionts and examined their response via metatranscriptome sequencing. Our data have uncovered genes with lineage-specific expression level adaptation in both host and photosymbiont taxa as well as the potential roles for bacterial associates. We have only scratched the surface of complex holobiont dynamics, yet these studies highlight the need for additional data and a comprehensive theoretical framework that integrates ecological and evolutionary time scales.

Speakers:

Mónica Medina

Pennsylvania State University

12:00 PM - 12:20 PM (EDT)

ODSS - Machine Reading for Precision Medicine

The advent of big data promises to revolutionize medicine by making it more personalized and effective, but big data also presents a grand challenge of information overload. For example, tumor sequencing has become routine in cancer treatment, yet interpreting the genomic data requires painstakingly curating knowledge from a vast biomedical literature, which grows by thousands of papers every day. Electronic medical records contain valuable information to speed up clinical trial recruitment and drug development, but curating such real-world evidence from clinical notes can take hours for a single patient. Natural language processing (NLP) can play a key role in interpreting big data for precision medicine. In particular, machine reading can help unlock knowledge from text by substantially improving curation efficiency. However, standard supervised methods require labeled examples, which are expensive and time-consuming to produce at scale. In this talk, I'll present Project Hanover, where we overcome the annotation bottleneck by combining deep learning with probabilistic logic, and by exploiting self-supervision from readily available resources such as ontologies and databases. This enables us to extract knowledge from millions of publications, reason efficiently with the resulting knowledge graph by learning neural embeddings of biomedical entities and relations, and apply the extracted knowledge and learned embeddings to supporting precision oncology.

Speakers:

Hoifung Poon

Microsoft Research

12:00 PM - 12:20 PM (EDT)

Function - CAFA4 overview

Speakers:

Damiano Piovesan

Department of Biomedical Sciences, University of Padova

Yisu Peng

Northeastern University

Predrag Radivojac

Northeastern University

Silvio Tosatto

Department of Biomedical Sciences, University of Padova

12:00 PM - 12:20 PM (EDT)

CAMDA - Effect of Tumor Purity on The Analysis of Gene Expression Data

The tumor microenvironment is comprised of tumor cells, stroma cells, immune cells, blood vessels, and other associated non-cancerous cells. Gene expression measurements on tumor samples are an average over cells in the microenvironment. However, research questions often seek answers about tumor cells rather than the surrounding non-tumor tissue. In this study, we investigate the effect of tumor purity – the proportion of tumor cells in a solid tumor sample – on the performance of two statistical methods used for analyzing gene expression data: differential network (DN) analysis and differential gene expression analysis (DGEA). We perform the DN analysis on a breast invasive carcinoma dataset. The results reveal cancer-progression related pathways when analyzing the high-purity samples, whereas many non-cancer-related pathways are obtained when analyzing the complete dataset. In addition, a simulation study is conducted to assess the effect of replacing low tumor purity samples compared to random sampling variability with two additional datasets: head and neck squamous cell carcinoma (HNSC) and lung squamous cell carcinoma (LUSC). The approach described in this study provides a general strategy for assessing the effect of tumor purity on any gene expression data analyses and can be applied to other types of cancers.

Speakers:

Seungjun Ahn

Department of Biostatistics, University of Florida

12:00 PM - 12:20 PM (EDT)

Education - HTrainDB: H3Africa Training Database

One of the main aims of the H3Africa Consortium is to improve its research capacity in genomics, bioinformatics and health in Africa. A number of training initiatives in the form of workshops, seminars and related activities is organized by the Consortium in view of strengthening its research capacity and training world-class trainers. Given the large amount of training activities and participants, it is imperative to maintain a centralized record of all pertinent information. The database has been created to monitor the training efforts and evaluate the effectiveness of this research capacity building initiative. HTrainDB has several functionalities including capturing several aspects of the consortium membership, registration for Consortium meetings, advertising and recording training activities, creating meeting polls, creating survey questionnaires and webforms for various purposes. It also lists the publications of all members and highlights their career timeline. It provides a centralized system to keep records related to the consortium membership, projects and activities. To this end it can capture aggregate data and showcase capacity building in genomics research in Africa. HTrainDB has more than 600 members and is exemplary to inform other research projects or consortia with large memberships on how to manage and archive research related administrative data.

Speakers:

Zahra Mungloo-Dilmohamud

University of Mauritius

12:00 PM - 12:20 PM (EDT)

MLCSB: Machine Learning - Causal network learning using a semi-supervised approach

Causal interplay between molecular components is central to the regulation of cellular behaviour. Data-driven learning of molecular network topology continues to be an active area of research, including development of methods that have a particular focus on learning causal relationships. These methods often use statistical models with explicit causal assumptions (for example, causal directed acyclic graphs). Here, we take an alternative approach and view causal network inference from a machine learning point of view. The idea is to allow direct learning of patterns of causal influence between nodes. Binary indicators of causal influence between pairs of variables are treated as "labels". Available data for the variables of interest are used to provide edge features for the labelling task. Background knowledge or any available interventional data provide labels on a subset of variable pairs. The task is to learn the remaining labels that are unobserved. Using a specific approach rooted in manifold semi-supervised learning we present empirical results on three different biological datasets, including data where causal effects can be verified experimentally. Our results demonstrate the efficacy and highly general nature of the approach as well as its simplicity from a user's point of view.

Speakers:

Steven Hill

MRC Biostatistics Unit, University of Cambridge

12:00 PM - 12:20 PM (EDT)

SysMod - Cellular robustness is not a byproduct of environmental flexibility

Cells show remarkable resilience against perturbations. Its evolutionary origin remains obscure. In order to leverage methods of systems biology to examine cellular robustness, a computationally accessible way of quantification is needed. Here, we present an unbiased metric of structural robustness in (genome-scale) metabolic models based on concepts prevalent in reliability engineering. The probability of failure (PoF) is defined as the (weighted) portion of all combinations of loss-of-function mutations that disable network functionality. It can be exactly determined if all essential reactions, all synthetic lethal pairs of reactions, all synthetic lethal triplets of reactions etc., are known. In theory, these minimal cut sets (MCSs) can be calculated for any network, but for large models the problem remains computationally intractable. We demonstrate that even at the genome-scale only the lowest-cardinality MCSs are required to efficiently approximate the PoF. We analyzed the robustness of 459 E. coli, Shigella, and Salmonella strains. In contrast to the congruence theory, which explains the origin of genetic robustness as a byproduct of selection for environmental flexibility, we found no correlation between robustness and the diversity of growth-supporting environments. On the contrary, our analysis indicates that the core-reactome, i.e. the set of reactions shared across strains, dominates robustness.

Speakers:

Jürgen Zanghellini

University of Vienna

12:20 PM - 12:40 PM (EDT)

Internet and Remote Sensing Data for Public Health Surveillance

In this talk, I will present opportunities and challenges of using data from Internet sources and remote sensing for public health surveillance. If properly mined and filtered, these data can be used to study the dynamics of disease propagation, risk factors and the relationship between human behavior and disease spread. This talk will include examples on the use of satellite images, social media postings and product reviews for surveillance of infectious diseases, food products and obesity prevalence.

Speakers:

ELAINE NSOESIE

BOSTON UNIVERSITY

12:20 PM - 12:30 PM (EDT)

Evolution and Comparative Genomics - A Probabilistic Framework for Cell Lineage Tree Reconstruction

Single-cell DNA sequencing (ScDNA-seq) technology enables a higher resolution look on the cells and has the potential to uncover the relationship between individual cells. ScDNA-seq is a fundamental tool for the evolutionary studies; however, it also introduces technological artefacts such as uneven coverage, allelic dropout, amplification and sequencing errors. In this study, we focus on human cells without copy number alterations. We present a Bayesian model based on scDNA-seq, that uncovers the difference between cells with the help of germline single nucleotide variations (gSNVs). The use of gSNVs as reference points enables us to accurately differentiate between somatic point mutations and the technological artefacts, especially the amplification errors. The model outputs a cell-to-cell distance matrix of each analysed pairs of loci, from which we reconstruct the cell lineage tree with bootstrapping and neighbour joining. We evaluate the reconstructed tree with transfer bootstrap expectation scores of branches and the Robinson-Foulds distance to the underlying tree structure. The model is embarrassingly parallel and with the use of dynamic programming and neighbour joining, we can analyse tens of thousands positions in the genome. The experiments showed high accuracy in tree reconstruction and the identification of subclones.

Speakers:

Hazal Koptagel

KTH Royal Institute of Technology, Science For Life Laboratory

12:20 PM - 12:40 PM (EDT)

Education - Staff Education to Accelerate the Cloud Adoption

The European Bioinformatics Institute (EMBL-EBI) is part of EMBL. We have a large number of staff developing databases and tools to host, analyse and share data and analytic results openly in the life sciences. Staff training to adopt new technologies such as cloud computing is very challenging. We have defined the long-term strategy to adopt cloud infrastructure to provide better access to scientists in Europe and around the world to accelerate their research with the diverse data stored, verified and visualised by EBI. ==SHORTENED TO SATISFY 200 WORD LIMIT. PLEASE SEE PDF FOR THE COMPLETE ABSTRACT== As a result, many research and service teams have launched their projects with the new skills in the clouds. This team has been invited to deliver cloud workshops for EU projects in person for EOSC-Life and remotely for BioExcel with additional projects in the near future. The Cloud Roundtable forum has now been extended to include participants from Wellcome Sanger Institute to foster closer cooperation in using cloud technologies with a new joint project proposed. With this systematic staff education, we have made our first step to lead EBI and EU projects to adopt cloud technologies.

Speakers:

Tony Wildish

European Bioinformatics Institute

12:20 PM - 12:40 PM (EDT)

MLCSB: Machine Learning - Factorized embeddings learns rich and biologically meaningful embedding spaces using factorized tensor decomposition

The recent development of sequencing technologies revolutionised our understanding of the inner workings of the cell as well as the way disease is treated. A single RNA sequencing (RNA-Seq) experiment, however, measures tens of thousands of parameters simultaneously. While the results are information rich, data analysis provides a challenge. Dimensionality reduction methods help with this task by extracting patterns from the data by compressing it into compact vector representations. We present the factorized embeddings (FE) model, a self-supervised deep learning algorithm that learns simultaneously, by tensor factorization, gene and sample representation spaces. We ran the model on RNA-Seq data from two large-scale cohorts and observed that the sample representation captures information on single-gene and global gene expression patterns. Moreover, we found that the gene representation space was organized such that tissue-specific genes, highly correlated genes as well as genes participating in the same GO terms were grouped. Finally, we compared the vector representation of samples learned by the FE model to other similar models on 49 regression tasks. We report that the FE-trained representations rank first or second in all of the tasks, surpassing, sometimes by a considerable margin, other representations.

Speakers:

Assya Trofimov

IRIC - Université de Montréal

12:20 PM - 12:40 PM (EDT)

SST03 - Metabolomic analysis of coral holobionts reveals markers of thermal stress

Coral reef systems are under threat globally due to anthropomorphic climate change. Understanding the response of the coral holobiont to abiotic and biotic stress is therefore crucial to aid conservation efforts. The most pressing problem is “coral bleaching”, usually precipitated by prolonged thermal stress that disrupts the algal symbiosis sustaining the holobiont. Here we used metabolomics to address the following questions: how does the coral holobiont metabolome respond to heat stress, and can we find diagnostic markers of that stress prior to the onset of bleaching? We studied the heat tolerant Montipora capitata and heat sensitive Pocillopora acuta coral species from the Hawaiian reef system. Untargeted LC-MS analysis uncovered a suite of known and novel metabolites that accumulate under heat stress, including a variety of dipeptides, some of which occupy hub positions in metabolite networks, and are present in both coral species. The structures of four of these were determined (Arginine-Glutamine, Alanine-Arginine, Valine-Arginine, Lysine-Glutamine). These dipeptides show differential accumulation in symbiotic and aposymbiotic (alga free) individuals of the sea anemone Aiptasia, suggesting their animal derivation and algal symbiont related function. We have identified a suite of metabolites associated with thermal stress in corals that provide tools to diagnose coral health in the wild.

Speakers:

Debashish Bhattacharya

Rutgers University

12:20 PM - 12:40 PM (EDT)

Function - CAFA4 Overview

Speakers:

Damiano Piovesan

Department of Biomedical Sciences, University of Padova

Yisu Peng

Northeastern University

Silvio Tosatto

Department of Biomedical Sciences, University of Padova

Predrag Radivojac

Northeastern University

12:20 PM - 12:40 PM (EDT)

SysMod - Closing remarks of the SysMod meeting 2020

A brief recap of the two meeting days and the SysMod community

Speakers:

Andreas Dräger

University of Tübingen, Institute for Bioinformatics and Medical Informatics (IBMI)

12:20 PM - 12:40 PM (EDT)

CAMDA - Mechanistic models of CMap drug perturbation functional profiles

We have used mechanistic models of pathway activities to generate complete catalogue of Cmap functional profiles that can be further used to detect reverse functional profiles in diseases, thus providing a functional basis and a potential biological interpretation for the drug-disease inverse matching.

Speakers:

Maria Peña-Chilet

CIBERER

12:30 PM - 12:40 PM (EDT)

Evolution and Comparative Genomics - Splicing-structure-based selection of protein isoforms improves the accuracy of gene tree reconstruction

Constructing accurate gene trees is important, as gene trees play a key role in several biological studies, such as species tree reconstruction, gene functional analysis and gene family evolution studies. Although several methods have provided large improvements in the construction and the correction of gene trees by making use of the relationship with a species tree in addition to multiple sequence alignments, there is still room for improvement on the accuracy of gene trees and the computing time. In particular, accounting for alternative splicing that allows eukaryote genes to produce multiple transcripts and proteins per gene is a way to improve the quality of multiple sequence alignments used to reconstruct gene trees.Current methods for gene tree reconstruction usually make use of a set of transcripts composed of one representative transcript per gene, to generate multiple sequence alignments which are then used to estimate gene trees. In this work, we present two new splicing-structure-based methods to estimate gene trees based on wisely selecting an accurate set of homologous transcripts based on their splicing structure to represent the genes of a gene family. The results show that the new methods compare favorably with the currently most used gene tree construction methods.

Speakers:

Esaie Kamela

Université de Sherbrooke

12:40 PM - 01:00 PM (EDT)

Evolution and Comparative Genomics - Inference of Population Admixture Network from Local Gene Genealogies: a Coalescent-based Maximum Likelihood Approach

Population admixture is an important subject in population genetics. Inferring population demographic history with admixture under the so-called admixture network model from population genetic data is an established problem in genetics. Existing admixture network inference approaches work with single genetic polymorphisms. While these methods are usually very fast, they don't fully utilize the information (e.g., linkage disequilibrium or LD) contained in population genetic data. In this paper, we develop a new admixture network inference method called GTmix. Different from existing methods, GTmix works with local gene genealogies that can be inferred from population haplotypes. Local gene genealogies represent the evolutionary history of sampled alleles and contain the LD information. GTmix performs coalescent-based maximum likelihood inference of admixture networks with inferred genealogies based on the well-known multispecies coalescent (MSC) model. GTmix utilizes various techniques to speed up likelihood computation on the MSC model and optimal network search. Our simulations show that GTmix can infer more accurate admixture networks with much smaller data than existing methods, even when these existing methods are given much larger data. GTmix is reasonably efficient and can analyze genetic datasets of current interests.

Speakers:

Yufeng Wu

Computer Science and Engineering Department, University of Connecticut

01:15 PM - 02:00 PM (EDT)

Lunch & Learn - F1000Research: Innovative Open Access publishing for software tools and beyond

Speakers:

Vicky Hellon

F1000Research

02:00 PM - 02:20 PM (EDT)

SCANGEN - Deep Representation Learning for Problems in Biology

Speakers:

Smita Krishnaswamy

Yale University

02:00 PM - 02:20 PM (EDT)

Education - IMPLEMENTATION OF ADVANCED BIOINFORMATICS EDUCATION ACROSS AFRICA

With more microbiome studies conducted by African based research groups, there is an increasing demand for knowledge in the design and analysis of microbiome studies and data, but high-quality bioinformatics courses are often hampered by factors such as lack of computational infrastructure and local expertise, among others. To address this need, H3ABioNet developed an intermediate 16S rRNA analysis course alongside experienced microbiome researchers who identified key topics ranging from designing microbiome studies to more practical topics such as introductory high-performance computing, microbiome analysis pipelines and downstream analyses conducted in R. Tools used in the course were packaged in Singularity containers to remove the overhead of installing individual tools, versions and libraries and a separate container was created for downstream analysis of results using Rstudio. The pulling, running and testing of the containers, software and analysis on various clusters was performed prior to the start of the course by all hosting classrooms. The pilot ran successfully in 2019 across 23 sites registered in 11 African countries, with more than 200 participants formally enrolled. It provides a model for delivering topic specific bioinformatics courses across Africa which overcome barriers such as unequal infrastructures, geographical distance, access to expertise and educational materials.

Speakers:

Verena Ras

H3ABioNet; University of Cape Town

02:00 PM - 03:00 PM (EDT)

CAMDA - Panel Presentations and Discussion

Join us for Panel Presentations and a Discussion session in the CAMDA Cafe room by clicking the CAMDA logo under "Cafe Connect". We will discuss latest insights and shape future Camda Contest Challenges!

Camda 2000-2020 - historical overview of previous CAMDA Conferences: Joaquin Dopazo, Clinical Bioinformatics Area, Fundación Progreso y Salud, Spain
Perspectives 2021 - next year's CAMDA challenges: Wenzhong Xiao, Stanford University & Massachusetts General Hospital, U.S.A.

Speakers:

Wenzhong Xiao

Joaquin Dopazo

Clinical Bioinformatics Area. Fundación Progreso y Salud, Sevilla

02:00 PM - 02:40 PM (EDT)

SST05 - HUBMAP Session Keynote: HuBMAP Data Collection Plans

Speakers:

Michael Snyder

Stanford University

02:00 PM - 02:20 PM (EDT)

Evolution and Comparative Genomics - EvoLSTM: Context-dependent models of sequence evolution using a sequence-to-sequence LSTM

Motivation: Accurate probabilistic models of sequence evolution are essential for a wide variety of bioinformatics tasks, including sequence alignment and phylogenetic inference. The ability to realistically simulate the evolution of sequence is also at the core of many benchmarking strategies. Yet mutational processes have complex context dependencies that remain poorly modeled and understood. Results: We introduce EvoLSTM, a recurrent neural network-based evolution simulator that captures mutational context dependencies. EvoLSTM uses a sequence-to-sequence LSTM model trained to predict mutation probabilities at each position of a given descendant ancestral sequence, taking into consideration the 14 flanking nucleotides. EvoLSTM can realistically simulate primate DNA sequence and reveals unexpectedly strong long-range context dependencies in mutation rates. Conclusion: EvoLSTM brings modern machine learning approaches to bear on sequence evolution. It will serve as a useful tool to study and simulate complex mutational processes.

Speakers:

Dongjoon Lim

McGill University

02:00 PM - 02:20 PM (EDT)

MLCSB: Machine Learning - Adversarial Deconfounding Autoencoder for Learning Robust Gene Expression Embeddings

The increasing number of gene expression profiles has enabled the use of complex models, such as deep unsupervised neural networks, to extract a latent space from these profiles. However, expression profiles, especially when collected in large numbers, inherently contain variations introduced by technical artifacts (e.g., batch effects) and uninteresting biological variables (e.g., age) in addition to the true signals of interest. These sources of variation, called confounders, produce embeddings that fail to capture biological variables of interest and transfer to different domains. To remedy this problem, we attempt to disentangle confounders from true signals to generate biologically informative embeddings by introducing the AD-AE (Adversarial Deconfounding AutoEncoder) approach. The AD-AE model consists of an autoencoder and an adversary network that we jointly train to generate embeddings that can encode as much information as possible without encoding any confounding signal. By applying AD-AE to two distinct gene expression datasets, we show that our model can (1) generate embeddings that do not encode confounder information, (2) conserve the biological signals present in the original space, and (3) generalize successfully across different confounder domains. We believe that this adversarial deconfounding approach can be the key to discovering robust expression patterns.

Speakers:

Ayse Dincer

Univeristy of Washington

02:00 PM - 02:40 PM (EDT)

iRNA - iRNA Keynote: Non-canonical base pair interactions improve the scalability and accuracy of the prediction and analysis of RNA 3D structures

A vast and complex network of base interactions stabilizes the 3D architecture of RNAs. Beyond the canonical Watson-Crick and Wobble base pairs, each pair of nucleotides can interact in up to 12 different ways. The frequency of each type of base pair varies a lot, but in most cases their occurrence is essential to shape the local or global geometry of structured RNAs. Non-canonical base pairs are primarily found within or between unpaired regions of the secondary structure commonly refereed as loops. Their concentration in loops creates sophisticated base interaction patterns that are representative of the 3D structure supported by this network. In this talk, we show that working with a graphical model based on the set of canonical and non-canonical base pairs offers novel opportunities to develop efficient algorithms for predicting and analyzing physical and biochemical properties of RNA 3D structures. We present applications of this framework to (i) the automated discovery of structural motifs in databases, (ii) the prediction of small molecules binding RNAs, and (ii) the prediction of the 3D structure of large RNAs. These results suggest novel avenues to study of the evolution of RNAs or accelerate RNA drug discovery.

Speakers:

Jérôme Waldispühl

McGill University

02:00 PM - 02:20 PM (EDT)

Lessons from Archives: Strategies for Collecting Sociocultural Data in Machine Learning

A growing body of work shows that many problems in fairness, accountability, transparency, and ethics in machine learning systems are rooted in decisions surrounding the data collection and annotation process. We argue that a new specialization should be formed within machine learning that is focused on methodologies for data collection and annotation: efforts that require institutional frameworks and procedures. Specifically for sociocultural data, parallels can be drawn from archives and libraries. Archives are the longest standing communal effort to gather human information and archive scholars have already developed the language and procedures to address and discuss many challenges pertaining to data collection such as consent, power, inclusivity, transparency, and ethics privacy. We discuss these five key approaches in document collection practices in archives that can inform data collection in sociocultural machine learning.

Speakers:

Timnit Gebru

Google Research

02:00 PM - 02:20 PM (EDT)

CAFA4 Talk 1

Speakers:

Shanfeng Zhu

Fudan University

02:20 PM - 02:40 PM (EDT)

CAFA4 Talk 2

Speakers:

Brigham Young University

02:40 PM - 03:00 PM (EDT)

Ethics, Bias, and the Adoption of AI in Biomedicine

Artificial intelligence (AI) has vast potential to revolutionize biomedicine across the translational science spectrum, from basic science to clinical research to practice and policymaking. However, recent high-profile cases – lawsuits alleging that patients’ privacy rights were violated when health data were shared with a technology company, errant treatment recommendations, and racially biased algorithms – suggest that addressing the ethical questions AI raises will be critical to its success. What does privacy require? Who owns data? Where is the line between informing choice through analytics, and manipulating it? Should some domains of the human experience be “off limits” for AI? How should the potential for biases and stigmatization as a result of AI applications be managed? Using a highly publicized real-world case of a biased resource allocation algorithm as an example, this presentation will describe three considerations that are critical for understanding how to manage ethical questions surrounding AI: context (i.e., how we manage bias and other ethical issues depends upon the data type and the circumstances under which data are collected and used); upstream action (i.e., bias and other ethical issues must be addressed at the earliest stages of AI development, not after the fact); and engagement (i.e., meaningfully engaging diverse stakeholders is essential to managing ethical issues). For AI to be most effective it must be ethical. Moving forward, there is a great need for empirical research into how ethical issues and their management affect the diffusion of AI-based innovations.

Speakers:

Matthew DeCamp

University of Colorado

02:40 PM - 03:00 PM (EDT)

CAFA4 Talk 3

Speakers:

Hongryul Ahn

The university of suwon

02:40 PM - 02:50 PM (EDT)

Evolution and Comparative Genomics - Highly-regulated and diverse NTP-based biological conflict systems with implications for emergence of multicellularity

Multicellular organizations are prone to infections even if a single cell is infected. We reveal novel highly-regulated chaperone-based systems that are likely used as survival tactic by prokaryotes with complex lifecycles. These architecturally-analogous systems have constant core modules coupled with highly-variable effector modules which is reminiscent of known biological conflict systems and co-evolutionary arms-race. The constant component is either an ATPase/GTPase and/or a peptidase that is activated in response to an invasive entity and causes effector deployment, that is additionally regulated by proteolytic processing or binding of nucleotide-derived signal. A third component senses invasive entities and transmits the signal. Effectors either: target invasive nucleic-acids or proteins; are inactive counterparts of host proteins that mediate decoy interactions with invasive molecules; or form macromolecular assemblages to cause host cell-death or containment of invasive entity. These apoptotic and immunity properties displayed by systems in phylogenetically-disparate multicellular prokaryotes are suggestive of evolutionary convergence for kin viability in multicellular organizations. Comparable protein domains appear to have organized into systems based on common principles in eukaryotic apoptosis. Thus, a similar operational “grammar” and shared “vocabulary” of protein domains in sensing and limiting infections during the multiple emergences of multicellularity across the tree of life is seen.

Speakers:

Gurmeet Kaur

NCBI, NML, NIH

02:40 PM - 02:50 PM (EDT)

iRNA - RNA Secondary Structure Prediction By Learning Unrolled Algorithms

RNA secondary structure prediction is one of the oldest computational problems in bioinformatics, which has been studied for more than 40 years. Usually, researchers tend to utilize dynamic programming to resolve it, which can be relatively slow, with the F1 score being around 0.6 and having difficulty in handling pseudoknots. In this paper, we address it from an entirely new angle, viewing it as a translation with constraints problem. We propose a novel end-to-end deep learning model, called E2Efold, which has the problem-specific constraints embedded in the network architecture. The core idea of E2Efold is to predict the RNA base-pairing matrix directly, and use an unrolled algorithm for constrained programming as the template for deep architectures to enforce constraints. With comprehensive experiments on benchmark datasets, we demonstrate the superior performance of E2Efold: it predicts significantly better structures compared to the previous state-of-the-art methods (especially for pseudoknotted structures), with the F1 score being around 0.8, while being as efficient as the fastest algorithm in terms of inference time. The original paper has been published as an oral paper in ICLR 2020. The code of E2Efold is available at https://github.com/ml4bio/e2efold.

Speakers:

Yu Li

KAUST

02:40 PM - 03:00 PM (EDT)

Education - Applying best practices to enhance bioinformatics training in Switzerland

SIB Swiss Institute of Bioinformatics training courses are created by following a cycle of best practices in training, which have been defined by the trainers’ communities in ISCB, GOBLET, ELIXIR and the SIB Training group. Trainers from the SIB Training group are encouraged to attend Train the trainer courses. As a consequence, SIB courses are clearly defined with specific learning objectives, target audiences, prerequisites, and active learning activities. Course pages are described with Bioschemas specifications and metadata, which provide valuable information to enable trainees to assess whether the courses meet their needs and background knowledge, a step towards FAIR training. Quality and impact metrics, together with training needs, are continuously collected, providing indications as to whether courses and the annual program need to be adapted. The integration of these best practices, together with the extensive expertise in bioinformatics from the SIB scientists, has resulted in courses that are consistently well evaluated and appreciated by PhD students, postdocs, and their PIs.

Speakers:

Diana Marek

SIB Swiss Institute of Bioinformatics

02:40 PM - 03:00 PM (EDT)

MLCSB: Machine Learning - SCIM: Universal Single-Cell Matching with Unpaired Feature Sets

Multi-modal molecular profiling of samples on a single-cell level can yield deeper insights into tissue microenvironment and disease dynamics. Profiling technologies like scRNA-seq often consume the analyzed cells and cellular correspondences between data modalities are lost. To exploit single cell ’omics technologies jointly, we propose Single-Cell data Integration via Matching (SCIM), a scalable and accurate approach to recover such correspondences in two or more technologies, even in the absence of overlapping feature sets. SCIM assumes that cells share a common underlying structure and reconstructs such technology-invariant latent space using an auto-encoder framework with an adversarial objective. Cell pairs across technologies are then identified using a customized bipartite matching scheme operating on the latent representations. We evaluate SCIM on a simulated branching process designed for scRNA-seq data (total of 192,000 cells) and show that the cell-to-cell matches reflect the same pseudotime (Pearson’s coefficient: 0.86). Moreover, we apply our method to a real-world melanoma tumor sample, and achieve 93% cell-matching accuracy with respect to cell-type label when aligning scRNA-seq and CyTOF datasets. SCIM is a scalable and flexible algorithm that bridges the gap between generation and integrative interpretation of diverse multi-modal data.

Speakers:

Stefan Stark

ETH Zürich

02:40 PM - 03:00 PM (EDT)

SCANGEN - Inferring the origins of pediatric brain tumors by single-cell analysis of the normal developing brain

Speakers:

Claudia Kleinman

03:20 PM - 03:40 PM (EDT)

MLCSB: Machine Learning - TinGa: fast and flexible trajectory inference with Growing Neural Gas

Motivation: During the last decade, trajectory inference methods have emerged as a novel framework to model cell developmental dynamics, most notably in the area of single-cell transcriptomics. At present, more than 70 trajectory inference methods have been published, and recent benchmarks showed that, while some methods perform well for certain trajectory types, overall there is still a lot of room for improvement. Results: In this work we present TinGa, a new trajectory inference model that is fast and flexible, and that is based on growing neural graphs. This allows TinGa to model both the most simple as well as most complex trajectory types. We performed an extensive comparison of TinGa to the five best existing methods for trajectory inference on a set of 250 datasets, including both synthetic as well as real datasets. Overall, TinGa obtained better results than all other methods on all ranges of data complexity, from the simplest linear datasets to the most complex disconnected graphs. In addition, TinGa obtained the fastest run times, showing that our method is thus one of the most versatile methods up to date. Availability: R scripts for running TinGa, comparing it to top existing methods and generating the figures of this paper are available at https://github.com/Helena-todd/researchgng

Speakers:

Helena Todorov

Ghent University

03:20 PM - 03:40 PM (EDT)

Evolution and Comparative Genomics - Copy Number Evolution with Weighted Aberrations in Cancer

Motivation: Copy number aberrations (CNAs), which delete or amplify large contiguous segments of the genome, are a common type of somatic mutation in cancer. Copy number profiles, representing the number of copies of each region of a genome, are readily obtained from whole-genome sequencing or microarrays. However, modeling copy number evolution is a substantial challenge, since CNAs alter contiguous segments of the genome and different CNAs may overlap with one another. A recent popular model for copy number evolution is the Copy Number Distance (CND), defined as the length of a shortest sequence of deletions and amplifications of contiguous segments that transforms one profile into the other. All events contribute equally to the CND; however, CNAs are observed to occur at different rates according to their length or genomic position and also vary across cancer type. Results: We introduce a weighted copy number distance that allows events to have varying weights, or probabilities, based on their length, position and type. We derive an efficient algorithm to compute the weighted copy number distance as well as the associated transformation, based on the observation that the constraint matrix of the underlying optimization problem is totally unimodular. We demonstrate the utility of the weighted copy number distance by showing that the weighted CND: improves phylogenetic reconstruction on simulated data where copy number aberrations occur with varying probabilities; aids in the derivation of phylogenies from ultra low-coverage single-cell DNA sequencing data; helps estimate CNA rates in a large pan-cancer dataset.

Speakers:

Ron Zeira

Princeton University

03:20 PM - 03:40 PM (EDT)

SCANGEN - Copy number aberrations from single-cell sequencing

Speakers:

Benjamin Raphael

Princeton University

03:20 PM - 03:45 PM (EDT)

SST05 - Infrastructure for Storing Massive Biological Data Sets

Speakers:

Nick Nystrom

Pittsburgh Supercomputing Center

03:20 PM - 04:00 PM (EDT)

CAMDA - Prediction of Drug Induced Liver Injury with different data sets and different end points

Motivation: Drug-induced liver injury (DILI) is one of the primary problems in drug development. Early prediction of DILI, would bring a significant reduction in the cost of clinical trials and faster development of drugs. Current study is aims at building predictive models of DILI potential of chemical compounds. Methods: We build predictive models for several alternative splits of compounds between DILI and non-DILI classes, using supervised Machine Learning algorithms. To this end we use chemical properties of the compounds under scrutiny, their effects on gene expression levels in 6 human cell-lines treated with them and their toxicological profiles. We first identity the most informative variables and then use them to build ML models. Individual models built using gene expression of single cell lines, chemical properties of compounds, their toxicology profiles are then combined using Super Learner approach. Results: We have obtained weakly predictive model for using molecular descriptors and DILI statistics, with AUC exceeding 0.7 for some DILI definitions. With one exception, gene expression profiles of human cell lines resulted were non-informative and resulted in random models. Gene expression profiles of HEPG2 cell line lead to statistically significant models (AUC=0.67) only for one definition of DILI.

Speakers:

Witold Rudnicki

University of Bialystok

03:20 PM - 03:40 PM (EDT)

Function - Unsupervised protein embeddings outperform hand-crafted sequence and structure features at predicting molecular function

Protein function prediction is a difficult bioinformatics problem. Many recent methods use deep neural networks to learn complex sequence representations and predict function from these. Deep supervised models require a lot of labeled training data which are not available for this task. However, a very large amount of protein sequences without functional labels is available. In this work we applied an existing deep sequence model that had been pre-trained in an unsupervised setting on the supervised task of protein function prediction. We found that this complex feature representation is effective for this task, outperforming hand-crafted features such as one-hot encoding of amino acids, k-mer counts, secondary structure and backbone angles. Also, it partly negates the need for deep prediction models, as a two-layer perceptron was enough to achieve state-of-the-art performance in the third Critical Assessment of Functional Annotation benchmark. We also show that combining this sequence representation with protein 3D structure information does not lead to performance improvement, hinting that three-dimensional structure is also potentially learned during the unsupervised pre-training.

Speakers:

Amelia Villegas-Morcillo

University of Granada

03:40 PM - 04:00 PM (EDT)

Function - Embeddings allow GO annotation transfer beyond homology

Understanding protein function is crucial for molecular and medical biology, nevertheless Gene Ontology (GO) annotations have manually been confirmed for fewer than 0.5% of all known protein sequences. Computational methods bridge this sequence-function gap, but the best prediction methods need evolutionary information to predict function. Here, we proposed a new method predicting GO terms through annotation transfer not using sequence similarity. Instead, the method uses SeqVec embeddings to transfer annotations between proteins through proximity in embedding space. SeqVec’s data driven feature extraction transferred knowledge from large unlabeled databases to smaller but labelled datasets (transfer learning). Replicating the conditions of CAFA3, our method reached an Fmax of 50%, 59%, and 65% for BPO, MFO, and CCO, respectively. This was numerically higher than all methods that had actually participated in CAFA3 for BPO and CCO and scored second for MFO. Restricting the lookup dataset to proteins with less than 20% pairwise sequence identity to the targets, performance dropped clearly (Fmax BPO 38%, MFO 46%, CCO 56%), but continued to clearly outperform simple homology-based inference. Thereby, the new method may help in annotating novel proteins not belonging to large families.

Speakers:

Maria Littmann

Technical University of Munich (TUM), Department of Informatics

03:40 PM - 04:00 PM (EDT)

"Moneyball" for Team Science? Fostering a Collaborative Workforce

Speakers:

Kara Hall

NCI

03:40 PM - 04:00 PM (EDT)

iRNA - Graph neural representational learning of RNA secondary structures for predicting RNA-protein interactions

Motivation: RNA-protein interactions are key effectors of post-transcriptional regulation. Significant experimental and bioinformatics efforts have been expended on characterizing protein binding mechanisms on the molecular level, and on highlighting the sequence and structural traits of RNA that impact the binding specificity for different proteins. Yet our ability to predict these interactions in silico remains relatively poor. Results: In this study, we introduce RPI-Net, a graph neural network approach for RNA-protein interaction prediction. RPI-Net learns and exploits a graph representation of RNA molecules, yielding significant performance gains over existing state-of-the-art approaches. We also introduce an approach to rectify particular type of sequence bias present in many CLIP-Seq data sets, and we show that correcting this bias is essential in order to learn meaningful predictors and properly evaluate their accuracy. Finally, we provide new approaches to interpret the trained models and extract simple, biologically-interpretable representations of the learned sequence and structural motifs.

Speakers:

Zichao Yan

McGill University

03:40 PM - 04:00 PM (EDT)

MLCSB: Machine Learning - Unsupervised Topological Alignment for Single-Cell Multi-Omics Integration

Motivation: Single-cell multi-omics data provide a comprehensive molecular view of cells. However, single- cell multi-omics datasets consist of unpaired cells measured with distinct unmatched features across modalities, making data integration challenging. Results: In this study, we present a novel algorithm, termed UnionCom, for the unsupervised topological alignment of single-cell multi-omics integration. UnionCom does not require any correspondence information, either among cells or among features. It first embeds the intrinsic low-dimensional structure of each single-cell dataset into a distance matrix of cells within the same dataset and then aligns the cells across single-cell multi-omics datasets by matching the distance matrices via a matrix optimization method. Finally, it projects the distinct unmatched features across single-cell datasets into a common embedding space for feature comparability of the aligned cells. To match the complex nonlinear geometrical distorted low-dimensional structures across datasets, UnionCom proposes and adjusts a global scaling parameter on distance matrices for aligning similar topological structures. It does not require one-to-one correspondence among cells across datasets, and it can accommodate samples with dataset-specific cell types. UnionCom outperforms state-of-the-art methods on both simulated and real single-cell multi-omics datasets. UnionCom is robust to parameter choices, as well as subsampling of features. Availability: UnionCom software is available at https://github.com/caokai1073/UnionCom.

Speakers:

Kai Cao

Academy of Mathematics and Systems Science, Chinese Academy of Sciences

03:40 PM - 04:00 PM (EDT)

SCANGEN - Joint Inference of Clonal Structure using Single-cell DNA-Seq and RNA-Seq data

Latest high-throughput single-cell RNA-sequencing (scRNA-seq) and DNA-sequencing (scDNA-seq) technologies enabled cell-resolved investigation of pathological tissue clones. However, it is still technically challenging to simultaneously measure the genome and transcriptome content of a single cell. In this work, we developed CCNMF – a new computational tool utilizing the Coupled-Clone Non-negative Matrix Factorization technique to jointly infer clonal structures in single-cell genomics and transcriptomics data. We benchmarked CCNMF using both simulated and real cell mixture derived datasets and fully demonstrated its robustness and accuracy. We also applied CCNMF to the paired scRNA and scDNA data from a triple-negative breast cancer xenograft, resolved its underlying clonal structures, and identified differential genes between cell clusters. In summary, CCNMF presents a joint and coherent approach to resolve the clonal genome and transcriptome structures, which will facilitate a better understanding of the cellular and tissue changes associated with disease development.

Speakers:

Xiangqi Bai

Academy of Mathematics and Systems Science, Chinese Academy of Sciences

03:40 PM - 04:00 PM (EDT)

Education - Embedding skills for a new profession by teaching programming in an immersive and authentic environment

Clinical Bioinformatics combines computer science with genomics in clinical practice. Trained clinical bioinformaticians are in short supply necessitating creative and flexibly-delivered education to fill the skills-gap. Our Introduction to Programming unit launched in 2019 as part of a PG-Cert teaching the fundamentals of Clinical Bioinformatics to a diverse cohort of distance learners. The unit simulated real-world experiences by building a situated learning environment that used agile project methods and authentic problem-solving activities to emulate clinical programming best-practice. Clear instructions, signposting and support, ensured comfort with course materials delivered using Jupyter Notebooks via GitHub. This use of industry-standard platforms and downloadable content also encouraged post-course lifelong-learning. The unit followed a social constructivist model geared to help students to learn individually and as a team. Synchronous online support from experienced facilitators helped encourage group-based peer-to-peer support which afforded more time for educators to support struggling students. By prioritising pedagogy over technology, the learning design resulted in incremental coding activities supporting a variety of learners. Methods such as Sprints provided real-world problem-based learning using real user-stories. The students directly contributed to the clinical bioinformatics toolkit by developing resources for personal practice or for co-development of the VariantValidator software, used in clinical practice worldwide.

Speakers:

Peter Causey-Freeman

University of Manchester

03:40 PM - 03:50 PM (EDT)

Evolution and Comparative Genomics - Reconstructing Tumor Evolutionary Histories and Clone Trees in Polynomial-time with SubMARine

Tumors contain multiple subpopulations of genetically distinct cancer cells. Reconstructing their evolutionary history can improve our understanding of how cancers develop and respond to treatment. Subclonal reconstruction methods infer the ancestral relationships among the subpopulations by constructing a clone tree. However, often multiple clone trees are consistent with the data. Current methods do not effectively characterize this uncertainty, and cannot scale to cancers with many subclonal populations. In this work we introduce a partial clone tree that defines a subset of the pairwise ancestral relationships in a clone tree, thereby implicitly representing the set of all clone trees that have these defined relationships. Also, we define a special partial clone tree, the Maximally-Constrained Ancestral Reconstruction (MAR), which summarizes all clone trees fitting the input data equally well. We describe SubMARine, a polynomial-time algorithm producing the subMAR, which approximates the MAR with specific guarantees. We also extend SubMARine to work with subclonal copy number aberrations. We show, both on simulated and a real lung cancer dataset, that SubMARine runs in less than 70 seconds, and that the subMAR equals the MAR in > 99.9% of cases where only a single tree exists. SubMARine is available at https://github. com/morrislab/submarine.

Speakers:

Linda K. Sundermann

University of Toronto

03:45 PM - 04:10 PM (EDT)

SST05 - Tools and Pipelines for the Analysis and Integration of HuBMAP data

To reconstruct a 3D human map HuBMAP intends to profile large amounts of high throughput biological data spanning several different organs, modalities and technologies. In this talk I will discuss the guiding principals underlying the decisions for selecting specific pipelines. I will mention some challenges related to implanting these pipelines in the context of a large and diverse consortium and will show some preliminary results for using these methods to analyze genomics, proteomics and imaging data.

Speakers:

Ziv Bar-Joseph

Carnegie Mellon University

03:50 PM - 04:00 PM (EDT)

Evolution and Comparative Genomics - Structural and transcriptional variation linked to protracted human frontal cortex development

The human frontal cortex is unusually large compared with many other species. The expansion of the human frontal cortex is accompanied by both connectivity and transcriptional changes. Yet, the developmental origins generating variation in frontal cortex circuitry across species remain unresolved. Nineteen genes, which encode filaments, synapse, and voltage-gated channels are especially enriched in the supragranular layers of the human cerebral cortex, which suggests enhanced cortico-cortical projections emerging from layer III. We identify species differences in connections with the use of diffusion MR tractography as well as gene expression in adulthood and in development to identify developmental mechanisms generating variation in frontal cortical circuitry. We demonstrate that increased expression of supragranular-enriched genes in frontal cortex layer III is concomitant with an expansion in cortico-cortical pathways projecting within the frontal cortex in humans relative to mice. We also demonstrate that the growth of the frontal cortex white matter and transcriptional profiles of supragranular-enriched genes are protracted in humans relative to mice. The expansion of projections emerging from the human frontal cortex emerges by extending frontal cortical circuitry development. Integrating gene expression with neuroimaging level phenotypes is an effective strategy to assess deviations in developmental programs leading to species differences in connections.

Speakers:

Christine Charvet

Delaware State University

04:00 PM - 04:20 PM (EDT)

ODSS - Reflections and Lessons from 15 Years of Training Computational Biologists

It is clear that there is an increasing need for computational biologists in both academia and industry. We launched a joint doctoral program to train computational biologists between Carnegie Mellon University and University of Pittsburgh in 2005, and this has been a continually evolving process since then in tandem with the rapid developments in the field, and the increasing demand for workforce in the field. I will present the challenges we faced in our interdisciplinary, inter-institutional program, and the solutions we came up with. As a program that admits students from a wide variety of backgrounds and aims at providing training in a rapidly evolving field, it has been critically important to monitor student progress and success as a function of their background and offer customized training when needed in line with career opportunities. A major lesson learned is the importance of familiarizing with the field early in education, before graduate studies; another is to gauge the career landscape and proactively adapt the training program to current and future needs.

Speakers:

Ivet Bahar

Distinguished Professor and JK Vries Chair, Computational & Systems Biology Department, School of Medicine, University of Pittsburgh and Associate Director, UP Drug Discovery Institute

04:00 PM - 04:10 PM (EDT)

iRNA - Practical Guidance for Genome-Wide RNA:DNA Triple Helix Prediction

Long noncoding RNAs (lncRNAs) play a key role in many cellular processes including chromatin regulation. To modify chromatin, lncRNAs often interact with DNA in a sequence-specific manner forming RNA:DNA triple helices. Computational tools for triple helices search do not always provide genome-wide predictions of sufficient quality. Here, we used four human lncRNAs (MEG3, DACOR1, TERC and HOTAIR) and their experimentally determined binding regions for evaluating triple helix parameters used by Triplexator software. We find out that 10 nt as a minimum length, 20% as a maximum error-rate and a minimum G-content of 70% or 40% provide the highest accuracy of triple helices predictions in terms of area under the curve (AUC). Additionally, we combined triple helix prediction with the lncRNA secondary structure and demonstrated that consideration of only single-stranded fragments of lncRNA predicted by RNAplfold with 0.95 or 0.5 thresholds for probability of pairing can further improve DNA-RNA the quality of triplexes prediction, especially in MEG3 case. This improvement can be explained by the number and characteristics of DBDs - regions of lncRNA that form the majority of the triplexes, detected by TDF software.

Speakers:

Elena Matveishina

Faculty of Bioengineering and Bioinformatics, Lomonosov Moscow State University, Moscow, Russia

04:00 PM - 04:10 PM (EDT)

Evolution and Comparative Genomics - CoCoCoNet: Conserved and Comparative Co-expression Across a Diverse Set of Species

Co-expression analysis has provided insight into gene function in many organisms from Arabidopsis to Zebrafish. Comparison across species has the potential to enrich these results, for example, by prioritizing among candidate human disease genes based on their network properties, or by finding alternative model systems where their co-expression is conserved. Here, we present CoCoCoNet as a tool for identifying conserved gene modules and comparing co-expression networks. CoCoCoNet is a resource for both data and methods, providing gold-standard networks and sophisticated tools for on-the-fly comparative analyses across 14 species. We show how CoCoCoNet can be used in two use cases. In the first, we demonstrate deep conservation of a nucleolus gene module across very divergent organisms, and in the second, we show how the heterogeneity of autism mechanisms in humans can be broken down by functional group, and translated to model organisms. CoCoCoNet is free to use and available at https://milton.cshl.edu/CoCoCoNet/, providing users with convenient access to both data and methods for cross-species analyses, opening up a range of potential research questions relevant to evolution and comparative genomics.

Speakers:

John Lee

Cold Spring Harbor Laboratory

04:00 PM - 04:20 PM (EDT)

SCANGEN - Integrative analysis of breast cancer survival based on spatial features

Complex cancer progression, such as tumour growth and invasion, is known to be impacted by the immune system and its spatial interaction with tumours. Two recent studies of breast cancer using multiplexed ion beam imaging by time-of-flight (MIBI-TOF) and imaging mass cytometry (IMC) have revealed that single-cell heterogeneity within the spatial tumour-microenvironment context is closely associated with patient subtypes (Keren et al. 2018, Jackson et al. 2020). However, a complete understanding of the ability of different features generated from these state-of-the-art technologies, such as cell type composition and tumour spatial patterns, to predict patient clinical outcomes is still limited. To this end, we will introduce a procedure to identify informative spatial and non-spatial features and develop a corresponding model to predict patient survival outcomes. We investigate integration methods such as optimal transport to reconstruct the spatial pattern of additional antibodies not measured in the imagining data using information from CyTOF. This allows us to expand the feature space by characterising the spatial pattern of existing and imputed protein expression, individual cell type, and cell type interaction using spatial statistics. Finally, we generalize the predictive features from different experiments and evaluate them via patient survival prediction performance.

Speakers:

Yingxin Lin

The University of Sydney

04:00 PM - 04:20 PM (EDT)

CAMDA - Improving Deep Learning Performance on Prediction of Drug-Induced Liver Injury

Drug-induced liver injury (DILI) is an important safety issue in the field of drug development. The potential to accurately predict DILI from both in vivo and in vitro studies would be an added advantage in analyzing the drug potential to cause DILI because hepatotoxicity might not be evident at the beginning stages of development. The annual international conference on Critical Assessment of Massive Data Analysis (CAMDA) releases challenges each year to tackle big data problems in the life sciences; the CMap Drug Safety challenge focuses on the prediction of DILI. In previous years, deep learning has been utilized but to un-stellar performance. This year (2020), we seek to improve the performance of deep learning on this challenge by investigating different methods of preprocessing data and network architectures. Our current leading model is able to predict severe DILI more accurately than previous deep learning results.

Speakers:

Peter Tran

Saint Louis University

04:00 PM - 04:20 PM (EDT)

MLCSB: Machine Learning - scNym: Semi-supervised neural networks for single cell identity classification

Single cell genomics experiments can reveal the keystone cellular actors in complex tissues. However, annotating cell type and state identities for each molecular profile in these experiments remains an analytical bottleneck. Here, we present scNym, a semi-supervised neural network that learns to transfer cell identity annotations from one experiment to another. scNym uses consistency regularization and entropy minimization techniques in the semi-supervised MixMatch framework to take advantage of information in both the labeled and unlabeled datasets. In benchmark experiments, we show that scNym offers superior performance to baseline approaches in transferring cell identity annotations across experiments performed with different technologies or in distinct biological conditions. We show with ablation experiments that semi-supervision techniques improved both the performance and calibration of scNym models. We also show that scNym models are well-calibrated and interpretable with saliency methods, allowing for review of model decisions by domain experts.

Speakers:

Jacob Kimmel

04:20 PM - 04:40 PM (EDT)

iRNA - PEGASAS: A pathway-guided approach for analyzing pre-mRNA alternative splicing during cancer progression

Aberrant pre-mRNA alternative splicing (AS) is widespread in cancer, but the causes and consequences of AS dysregulation during cancer progression are not well understood. We developed a novel computational framework, PEGASAS, as a pathway-guided approach for examining the effects of oncogenic signaling on exon incorporation. PEGASAS was designed to study the interplay among oncogenic signaling, AS, and affected biological processes. In this study, we applied PEGASAS to define the AS landscape across prostate cancer disease states and the relationship between splicing and known driver alterations. We compiled a meta-dataset of RNA-seq data of 876 tissue samples from publicly available sources, covering a range of disease states, from normal tissues to aggressive metastatic tumors. PEGASAS analysis revealed a correlation between Myc signaling and splicing changes in RNA binding proteins (RBPs), suggestive of a previously undescribed auto-regulatory phenomenon. We experimentally verified this result in a human prostate cell transformation assay. Our findings establish a role for Myc in regulating RNA processing by controlling incorporation of nonsense mediated decay-determinant exons in RBP-encoding genes. In conclusion, PEGASAS can mine large-scale transcriptomic data to connect changes in pre-mRNA AS with oncogenic alterations that are common to many cancer types.

Speakers:

Yang Pan

UCLA

04:20 PM - 04:40 PM (EDT)

Evolution and Comparative Genomics - Phylogenetic double placement of mixed samples

Motivation: Consider a simple computational problem. The inputs are (i) the set of mixed reads generated from a sample that combines two organisms and (ii) separate sets of reads for several reference genomes of known origins. The goal is to find the two organisms that constitute the mixed sample. When constituents are absent from the reference set, we seek to phylogenetically position them with respect to the underlying tree of the reference species. This simple yet fundamental problem (which we call phylogenetic double-placement) has enjoyed surprisingly little attention in the literature. As genome skimming (low-pass sequencing of genomes at low coverage, precluding assembly) becomes more prevalent, this problem finds wide-ranging applications in areas as varied as biodiversity research, food production and provenance, and evolutionary reconstruction. Results: We introduce a model that relates distances between a mixed sample and reference species to the distances between constituents and reference species. Our model is based on Jaccard indices computed between each sample represented as k-mer sets. The model, built on several assumptions and approximations, allows us to formalize the phylogenetic double-placement problem as a non- convex optimization problem that deconvolutes mixture distances and performs phylogenetic placement simultaneously. Using a variety of techniques, we are able to solve this optimization problem numerically. We test the resulting method, called MISA, on a varied set of simulated and biological datasets. Despite all the assumptions used, the method performs remarkably well in practice. Availability: The sofware and data are available at https://github.com/balabanmetin/misa.

Speakers:

Metin Balaban

University of California San Diego

04:20 PM - 04:40 PM (EDT)

CAMDA - Gene expression signature-based machine learning classifier of drug-induced liver injury

Drug-induced liver injury (DILI) is considered a primary factor in regulatory clearance for drug development, and there is a pressing need to develop and evaluate new prediction models for DILI. The CAMDA 2020 CMap Drug Safety Challenge included 422 drugs for training and 195 drugs with blinded labels for testing to predict four types of DILI classes (DILI1, DILI3, DILI5, and DILI6). Our approach utilized the machine learning (ML) approach focusing on drug perturbation gene expression signatures from the six human cell lines (PHH, HEPG2, HA1E, A375, MCF7, and PC3). We created representative expression signatures, the 250 most up-regulated and 250 down-regulated genes, for each drug using Kruskal-Borda merging of ranked z-scores profiles. Various ML algorithms, including random forest (RF), recursive partitioning and regression trees (RPART), support vector machine (SVM), generalized linear model (GLM), and naïve-Bayes classifier were built and evaluated using 100 times 5-fold cross-validation. The initial model results range from a ROC value of 0.818 in the RF DILI3 to 0.491 in GLM DILI6. These models are still a work in progress, in which data from Tox21, FAERS, and Mold2 will need to be incorporated alongside the gene expression response garnered from CMap.

Speakers:

Junguk Hur

University of North Dakota

04:20 PM - 04:40 PM (EDT)

ODSS - Integrating Biomedical Informatics and Data Science to Prepare the Precision Medicine Workforce

As biomedicine, and in particular, translational research, have entered the era of big data and artificial intelligence, the Washington University School of Medicine has developed a precision medicine roadmap intended to respond to such trends. At the core of this roadmap is the promise of precise and data-driven approaches to enabling the delivery of the right treatment to the right patient at the right time, all with the objective of saving and improving lives. As part of our roadmap, we believe that the biomedical research teams of the future will increasingly need to utilize multi-disciplinary approaches, notably incorporating the use of biomedical informatics and data science theories and methods. Based upon this belief, we have launched a comprehensive portfolio of practical, in-career, and scientific training programs, all of which transcend traditional disciplinary boundaries, and that ultimately seek to create a workforce capable of delivering on the promise of precision medicine. In this presentation, we will review the structure, curriculum, and lessons learned as a result of establishing such education and workforce development programs. In particular, we will focus on the critical need to combine both biomedical informatics and data science methodologies with driving biological and clinical problems, so as to engage and support learners in a highly contextualized and experiential learning environment.

Speakers:

Philip Payne

Washington University Institute for Informatics

04:30 PM - 04:40 PM (EDT)

MLCSB: Machine Learning - A Non-Parametric Bayesian Framework for Detecting Coregulated Splicing Signals in Heterogeneous RNA Datasets with Applications to Acute Myeloid Leukemia

Analysis of RNASeq data from large patient cohorts can reveal transcriptomic perturbations that are associated with disease. This is typically framed as an unsupervised learning task to discover latent structure in a data matrix. However, the heterogeneity of these datasets makes such analysis challenging. For example, in acute myeloid leukemia, mutations in splice factor genes occurring in a subset of the patients may only result in alteration of a subset of splicing events. Thus, there is a need to identify “tiles”, defined by a subset of samples and splicing events with abnormal signals. Although algorithms exist for this task, they fail to model splicing data. To address these challenges, we propose CHESSBOARD, a non-parametric Bayesian model for unsupervised discovery of tiles. Our algorithm does not require a priori knowledge of the number of the tiles and uses a unique missing value model for cancer data. First, we apply our model to synthetic datasets and show it outperforms several baseline approaches. Next, we show that it recovers tiles characterized by splicing aberrations which are reproducible in multiple AML patient cohorts. Finally, we show that tiles we discover are correlated with drug response to therapeutics, pointing to translational potential of our findings.

Speakers:

David Wang

University of Pennsylvania

04:35 PM - 05:00 PM (EDT)

SST05 - Common coordinates for registering human scale data at multiple scales

The Common Coordinate System (CCF) consists of ontologies, reference object libraries, and computer software (e.g., user interfaces) that enable biomedical experts to semantically annotate tissue samples and to precisely describe their locations in the human body (“registration”), align multi-modal tissue data extracted from different individuals to a reference coordinate system (“mapping”) and, provide tools for searching and browsing HuBMAP data at multiple levels, from the whole body down to single cells (“exploration”).

Speakers:

Katy Borner

Indiana University

04:45 PM - 06:00 PM (EDT)

Panel Discussion

05:00 PM - 05:10 PM (EDT)

MLCSB: Machine Learning - Knowledge-primed neural networks enable biologically interpretable deep learning on single-cell sequencing data

Deep learning algorithms are powerful predictors but generally provide little insight into the functions that underlie a prediction. We hypothesized that deep learning on biological networks instead of artificial networks assigns meaningful weights that can be readily interpreted, thus developing knowledge-primed neural networks (KPNNs). In KPNNs, every node has a molecular equivalent (protein activity) and every edge has a mechanistic interpretation (a regulatory interaction). KPNNs thereby connect gene expression to cellular phenotypes through large regulatory networks. After training on large single-cell transcriptomics datasets, KPNNs reveal important regulatory proteins. We validate KPNN interpretability on simulated data and biological systems with known ground truth, and reveal regulatory proteins in underexplored systems. We show that KPNN interpretations largely differ from standard interpretation approaches that are focused on features, and that interpretations are lost in shuffled networks. These results are enabled by an optimized learning method that stabilizes node weights after random initiation and controls for connectivity biases in biological networks. In summary, KPNNs combine the predictive power of deep learning with the interpretability of biological networks. While demonstrated on single-cell sequencing data and regulatory networks, our method is broadly relevant to all research areas where domain knowledge can be represented as networks.

Speakers:

Nikolaus Fortelny

CeMM Center for Molecular Medicine of the Austrian Academy of Sciences

05:00 PM - 05:20 PM (EDT)

Education - Guidelines for curriculum and course development in higher education and training

Background: Curriculum and instructional development should follow a formal process. Although the focus in formal curriculum theory is on long-term programs of study, the process is also applicable to shorter-form Learning Experiences (LEs) (single courses, lessons, or training sessions). Successful curricula and instruction support learners as they develop from entry-level performance to the minimum qualification for completing a program or course, articulated in terms of Learning Outcomes (LOs). These considerations have been encapsulated in an iterative model of curriculum and instructional design, with guidelines for its use. Output and conclusion: The starting point is the articulation of target LOs: everything follows from these, including the selection of LEs, content, the development of assessments, and evaluation of the resulting curriculum/instruction. The iterative process can be used in curriculum and instructional development, and provide a set of practical guidelines for curriculum and course preparation. The essential features effective curriculum and instruction (i.e., that achieves its stated LOs for the majority of learners) is presented here, to offer practical guidance and support for devising and evaluating both short- and long-form teaching.

Speakers:

Rochelle E. Tractenberg

Georgetown University and the Collaborative for Research on Outcomes and -Metrics

05:00 PM - 05:20 PM (EDT)

Function - INGA protein function prediction for the dark proteome

Our current knowledge of complex biological systems is stored in a computable form through the Gene Ontology (GO) which provides a comprehensive description of gene function. Prediction of GO terms from the sequence remains, however, a challenging task, which is particularly critical for novel genomes. Here we present INGA version 3.0, a new version of the INGA software for protein function prediction. INGA exploits homology, domain architecture and information from the ‘dark proteome’, i.e. all those features difficult to detect looking at sequence conservation. In the new version, which was implemented for CAFA4, we also included low complexity, coiled coil and homorepeat information. Dark features are used to enrich domain architectures allowing to increase the specificity of related functions. Also, they allow the characterization of those proteins not matching any known domain. INGA was ranked in the top ten methods on CAFA2 and second for the majority of CAFA3 challenges. The new algorithm can process entire genomes in a few hours or even less when additional input files are provided. The INGA web server, databases and benchmarking are available from URL: https://inga.bio.unipd.it/.

Speakers:

Damiano Piovesan

Department of Biomedical Sciences, University of Padova

05:00 PM - 05:20 PM (EDT)

CAMDA - AITL: Adversarial Inductive Transfer Learning with input and output space adaptation for pharmacogenomics

Motivation: the goal of pharmacogenomics is to predict drug response in patients using their single- or multi-omics data. A major challenge is that clinical data (i.e. patients) with drug response outcome is very limited, creating a need for transfer learning to bridge the gap between large pre-clinical pharmacogenomics datasets (e.g. cancer cell lines), as a source domain, and clinical datasets as a target domain. Two major discrepancies exist between pre-clinical and clinical datasets: 1) in the input space, the gene expression data due to difference in the basic biology, and 2) in the output space, the different measures of the drug response. Therefore, training a computational model on cell lines and testing it on patients violates the i.i.d assumption that train and test data are from the same distribution. Results: We propose Adversarial Inductive Transfer Learning (AITL), a deep neural network method for addressing discrepancies in input and output space between the pre-clinical and clinical datasets. AITL takes gene expression of patients and cell lines as the input, employs adversarial domain adaptation and multi-task learning to address these discrepancies, and predicts the drug response as the output. To the best of our knowledge, AITL is the first adversarial inductive transfer learning method to address both input and output discrepancies. Experimental results indicate that AITL outperforms state-of-the-art pharmacogenomics and transfer learning baselines and may guide precision oncology more accurately.

Speakers:

Hossein Noghabi

Simon Fraser University

05:00 PM - 05:20 PM (EDT)

Evolution and Comparative Genomics - Sampling and Summarizing Transmission Trees with Multi-strain Infections

Motivation: The combination of genomic and epidemiological data hold the potential to enable accurate pathogen transmission history inference. However, the inference of outbreak transmission histories remains challenging due to various factors such as within-host pathogen diversity and multi-strain infections. Current computational methods ignore within-host diversity and/or multi-strain infections, often failing to accurately infer the transmission history. Thus, there is a need for efficient computational methods for transmission tree inference that accommodate the complexities of real data. Results: We formulate the Direct Transmission Inference (DTI) problem for inferring transmission trees that support multi-strain infections given a timed phylogeny and additional epidemiological data. We establish hardness for the decision and counting version of the DTI problem. We introduce TiTUS, a method that uses SATISFIABILITY to almost uniformly sample from the space of transmission trees. We introduce criteria that prioritizes parsimonious transmission trees that we subsequently summarize using a novel consensus tree approach. We demonstrate TiTUS’s ability to accurately reconstruct transmission trees on simulated data as well as a documented HIV transmission chain. Availability: https://github.com/elkebir-group/TiTUS

Speakers:

Palash Sashittal

University Of Illinois at Urbana-Champaign

05:00 PM - 05:20 PM (EDT)

SCANGEN - Single cell whole genome sequencing for studying cancer evolution

Speakers:

Sohrab Shah

Memorial Sloan Kettering Cancer Center

05:00 PM - 05:30 PM (EDT)

SST05 - How can I interact with, use, develop methods for and obtain funding to work with HuBMAP data?

Speakers:

Ajay Pillai

National Institutes of Health

Katy Borner

Indiana University

Ziv Bar-Joseph

Carnegie Mellon University

Nick Nystrom

Pittsburgh Supercomputing Center

Nils Gehlenborg

Harvard Medical School

Michael Snyder

Stanford University

05:00 PM - 05:40 PM (EDT)

iRNA - iRNA Keynote: Systematic approaches to study the subcellular localization properties of RNAs and RNA Binding Proteins

Our laboratory seeks to understand the biological functions and regulatory mechanisms of RNA intracellular localization. Our projects aim to elucidate the normal functions of RNA trafficking in the maintenance of genome stability and cell polarity, and how disruption of these pathways can contribute to the aetiology of diseases such as cancer and neuromuscular disorders. For this work, we combine the versatility of Drosophila genetics with high-throughput molecular imaging and functional genomics approaches in fly and human cellular models. During this presentation, I will provide an update in our efforts i) to assess the global cellular distribution properties of the human and fly transcriptomes using fractionation-sequencing methodologies, ii) to systematically map the cytotopic distribution properties of human RNA binding proteins, and iii) to characterize perturbations in post-transcriptional regulatory pathways that may be linked to disease aetiology.

Speakers:

Eric Lécuyer

IRCM

05:10 PM - 05:20 PM (EDT)

MLCSB: Machine Learning - DeepPLIER: a deep learning approach to pathway-level representation of gene expression data

Extracting latent representation of gene expression data can provide insights into components and activations of gene pathways and networks. Inspired by the Pathway-level information extractor (PLIER) method, we propose a deep learning model, DeepPLIER, that incorporates pathway information in its architecture as prior for extracting latent variables (LVs) from gene expression data. DeepPLIER constructs LVs as combinations of known pathways by using partially connected nodes, but also includes fully connected nodes to correct for pathway misspecifications and to learn potentially unknown pathways from data. By incorporating pathway information, extraction of biologically relevant LVs is encouraged. Using simulation, we show that DeepPLIER achieves higher LV estimation accuracy than PLIER, especially in scenarios where prior pathways are partially missing. Using two large gene expression datasets from bulk brain tissues (ROSMAP and CommonMind), we show that DeepPLIER attains higher LV replication than PLIER. As well, we show that some of the identified LVs represent cell type proportions, and these LVs more accurately align with experimental estimates from immunohistochemistry than PLIER for a number of brain cell types.

Speakers:

Yichen Zhang

05:30 PM - 05:40 PM (EDT)

Evolution and Comparative Genomics - Reading the book of life: the language of proteins

Background Genomes are remarkably similar to natural language texts. From an information theory perspective, we can think of amino acid residues as letters, protein domains as words, and proteins as sentences consisting of ordered arrangements of protein domains (domain architectures). This work describes our recent efforts towards understanding the linguistic properties of genomes. Results Our recent work showed that the complexity of “grammars” in all major branches of life is close to a universal constant of ~1.2 bits. This is remarkably similar to natural languages; such an--yet unexplained--universal information gain has been observed and generally used to determine whether a series of symbols represent a language. In this work, we describe the implications of this work and its extension in various areas with a particular emphasis on measuring the proteome complexities in human tissues. Conclusion Our work established the similarity between natural languages and genomes and showed, for the first time, that there exists a “quasi-universal grammar” of protein domains and measured the minimal complexity of proteome required for a functional cell. We also describe the proteome complexities in human tissues and their functional significance.

Speakers:

Malay Basu

University of Alabama, Birmingham

05:40 PM - 06:00 PM (EDT)

CAMDA - Closing remarks and awards

Speakers:

David Kreil

05:40 PM - 06:00 PM (EDT)

iRNA - iRNA: concluding Remarks

Wrap-up

Speakers:

Klemens Hertel

UC Irvine

Michelle Scott

Université de Sherbrooke

Yoseph Barash

University of Pennsylvania

05:40 PM - 06:00 PM (EDT)

Function - Protein function inference via curated aggregate co-expression networks.

Background We explore the specific utility of co-expression data for the CAFA challenge. Our lab has assembled a set of highly-curated aggregate co-expression networks derived from 895 datasets (~39 517 samples) from 14 species, providing a well-powered source for inference. Methods We tested performance via the standard CAFA approach of calculating Fmax per protein. Our “direct expression” method simply finds the top co-expressed genes of benchmark genes, and transfers functions where available. Our “orthoexpression” extends species coverage by first looking up top homologs to the benchmarks, finding coexpressed genes from those, and then transferring functions. Results/Conclusions We found evidence that expression data contains information that can be used to improve functional annotation. Both our methods performed well relative to the pure sequence-based (phmmer) baseline, particularly for BP. To quantify potential performance improvements we ask how well we could do if we knew in advance which method is better. Plots of predictions against one another show substantial room for improvement. We find that while aggregated coexpression data can successfully be used to improve protein function prediction, integration with sequence-based methods is a major area for potential progress in the field.

Speakers:

John Hover

Cold Spring Harbor Laboratory

05:40 PM - 05:50 PM (EDT)

Evolution and Comparative Genomics - What is the structure of the ‘evolutionary model space’ for proteins?

Estimates of amino acid exchangeabilities are central to to models of protein evolution; there have been many efforts to estimate those parameters using from large numbers of proteins. Although models trained in this way can be useful for phylogenetic analyses, they provide limited information about the process of protein evolution. Recent studies have revealed that patterns of protein evolution (assessed using amino acid exchangeability parameters) vary across the tree of life; in other words, the processes underlying protein evolution are non-homogeneous. However, optimizing a 20-state non-homogenous model requires the estimation of many. of free parameters. Thus, it represents a challenging computational problem. There are two straightforward ways to simplify this problem: 1) estimate parameters for the best-approximating time-reversible model using restricted taxon sets; or 2) reduce the number of free parameters by constraining the model using biochemical information. Different protein structural environments are also associated with distinct amino acid substitution patterns; this results in a mixture of underlying models that also differ among taxa. Efforts to use the approaches described above, in combination with structural partitioning, to understand the space of protein evolutionary models will be described.

Speakers:

Edward Braun

Univeristy of Florida

05:50 PM - 06:00 PM (EDT)

Evolution and Comparative Genomics - On quantifying evolutionary importance of protein sites: A tale of two measures

A key challenge in evolutionary biology is the quantification of selective pressure on proteins and other biological macromolecules at single-site resolution. The evolutionary importance of a protein site under purifying selection is typically measured by the degree of conservation of the protein site itself. A possible alternative measure is the strength of the site-induced conservation gradient in the rest of the protein structure. Here, we show that despite major differences, there is a linear relationship between the two measures such that more conserved protein sites also induce stronger conservation gradient in the protein structure. This linear relationship is universal as it holds for different types of proteins and functional sites. Our results show that generally, the selective pressure acting on a functional site percolates through the rest of the protein via residue-residue contacts. Surprisingly however, catalytic sites in enzymes are the principal exception to this rule. Catalytic sites induce significantly stronger conservation gradients in the rest of the protein than expected from the degree of conservation of the site alone. The uniquely stringent requirement for the active site to selectively stabilize the transition state of the catalyzed chemical reaction imposes additional selective constraints on the rest of the enzyme.

Speakers:

Avital Sharir-Ivry

McGill University

09:30 AM - 10:30 AM (EDT)

Keynotes - Machine learning for structural and functional genomics

Recent advances in network and functional genomics have enabled large-scale measurements of molecular interactions, functional activities, and the impact of genetic perturbations. Identifying connections, patterns, and deeper functional annotations from heterogeneous measurements will enhance our capability to predict protein function, discover their roles in biological processes underlying diseases, and develop novel therapeutics. In this talk, I will cover a few machine algorithms that interrogate molecular interactions, perturbation screens, structural data, and evolutionary information to understand protein functions. First, I will present a network-based representation learning algorithm that integrates multiple molecular interactomes into compact topological embeddings for protein function inference. I will describe its application to discovering new disease factors and subnetworks from genetic perturbations and variations. Finally, I will present a few deep learning algorithms for protein structure prediction and sequence-to-function mapping. Applications to protein engineering and disease-associated mutation prediction will be introduced.

Speakers:

Jian Peng

University Of Illinois at Urbana-Champaign

10:40 AM - 11:00 AM (EDT)

CompMS - Community standard for reporting the experimental design in proteomics experiments: From samples to data files

Sharing of proteomics data within the research community has been greatly facilitated by the development of open data standards such as mzIdentML, mzML, and mzTab. In addition, the ProteomeXchange consortium of proteomics resources has enabled the data submission and dissemination of public MS proteomics data worldwide. However, several authors have pointed out the lack of suitable experiment-related metadata in public proteomics datasets making data reuse by the community difficult. In particular, datasets contain limited sample information about the disease, tissue, cell type, and treatment, among others; and the relationship between the samples and the mass spectrometry files (RAW) is missing in most cases. Here, we present a community standard data model and file format to represent the experimental design for proteomics experiments. The proposed data model is based on the Sample and Data Relationship Format (SDRF), which originates from the metadata file format developed by the transcriptomics community called MAGE-TAB. The aim is to define a set of guidelines to support the annotation of the experimental design in proteomics datasets in the public domain, enabling future integration among multi-omics experiments.

Speakers:

Yasset Perez-Riverol

European Bioinformatics Institute (EMBL-EBI), Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SD, UK

10:40 AM - 10:45 AM (EDT)

RegSys - RegSys Introduction

Speakers:

Alejandra Medina-Rivera

International Laboratory for Human Genome Research, Universidad Nacional Autónoma de México, Querétato, México

Shaun Mahony

Center for Eukaryotic Gene Regulation, The Pennsylvania State University, University Park, PA

10:40 AM - 11:40 AM (EDT)

Hitseq Keynote: Genetic and epigenetic maps of human centromeric regions

Speakers:

Karen Miga

UCSC

10:40 AM - 10:50 AM (EDT)

BioVis - BioVis Opening

Speakers:

Cagatay Turkay

University of Warwick

10:40 AM - 11:00 AM (EDT)

3DSIG - Deep Learning of Protein Structural Classes: Any Evidence for an 'Urfold'?

Recent advances in protein structure determination and prediction offer new opportunities to decipher relationships amongst proteins--a task that entails 3D structure comparison and classification. Historically, protein domain classification has been somewhat manual and heuristic. While CATH and related resources represent significant steps towards a more systematic (and automatable) approach, more scalable and objective classification methods, e.g., grounded in machine learning, could be informative. Indeed, comparative analyses of protein structures via Deep Learning (DL), though it may entail large-scale restructuring of classification schemes, could uncover distant relationships. We have developed new DL models for domain structures (including physicochemical properties), focused initially at CATH's homologous superfamily (SF) level. Adopting DL approaches to image classification and segmentation, we have devised and applied a hybrid convolutional autoencoder architecture that allows SF-specific models to learn features that, in a sense, 'define' the various homologous SFs. We quantitatively evaluate pairwise 'distances' between SFs by building one model per SF and comparing the loss functions of the models. Clustering on these distance matrices provides a new view of protein interrelationships--a view that extends beyond simple structural/geometric similarity, towards the realm of structure/function properties, and that is consistent with a recently proposed 'Urfold' concept.

Speakers:

Eli Draizen

University of Virginia

10:40 AM - 11:20 AM (EDT)

NetBio: Network Biology - NetBio Keynote: Network Medicine: From Cellular Networks to the Human Diseasome

Given the functional interdependencies between the molecular components in a human cell, a disease is rarely a consequence of an abnormality in a single gene, but reflects the perturbations of the complex intracellular network. The emerging tools of network medicine offer a platform to explore systematically not only the molecular complexity of a particular disease, leading to the identification of disease modules and pathways, but also the molecular relationships between apparently distinct (patho) phenotypes. Advances in this direction are essential to identify new disease genes, to uncover the biological significance of disease-associated mutations identified by genome-wide association studies and full genome sequencing, and to identify drug targets and biomarkers for complex diseases.

Speakers:

Albert-Laszlo Barabasi

10:40 AM - 11:20 AM (EDT)

TransMed - Transmed Keynote: 20 Challenges of AI in Medicine

The opportunities for using artificial intelligence (AI) and machine learning to improve healthcare are endless. Recent successes include the use of deep learning to identify clinically important patterns in radiographic images. However, there are numerous important challenges which must be addressed before AI can become widely adopted and incorporated into clinical workflows. Some of these challenges come from healthcare data. For example, clinical data from electronic health records (EHR) are notoriously noisy and incomplete. Some of these challenges come from the limitations of AI algorithms. For example, each algorithm looks at data in a different manner. How do you know which are the right methods to employ for a given data set? We will review 10 important clinical data challenges and 10 AI challenges which can impede progress in this area. A number of specific examples and some possible solutions will be provided.

Speakers:

Jason Moore

University of Pennsylvania

10:40 AM - 11:20 AM (EDT)

MICROBIOME - Microbiome Keynote: Assembling and modelling complex microbiomes mediating host-pathogen interactions

3DSIG - EM Map Segmentation and De Novo Protein Structure Modeling for Multiple Chain Complexes with MAINMAST

The significant progress of cryo-electron microscopy (cryo-EM) poses a pressing need for software for structural interpretation of EM maps. Methods for map segmentation is particularly needed for the modeling because most of the modeling methods are designed for building a single protein structure. Here, we developed new software, MAINMASTseg, for segmenting maps with symmetry. Unlike existing segmentation methods that merely consider densities in an input EM map, MAINMASTseg captures underlying molecular structures by constructing a skeleton that connects local dense points in the map. MAINMASTseg performed significantly better than other popular existing methods.

Speakers:

Genki Terashi

Department of Biological Sciences, Purdue University

11:10 AM - 11:20 AM (EDT)

3DSIG - Protein Contact Map De-noising Using Generative Adversarial Networks

Protein residue-residue contact prediction from protein sequence information has made substantial improvement in the past years and has been a driving force in the protein structure prediction field. In this work, we propose a novel contact map denoising method, ContactGAN, which uses Generative adversarial networks (GAN). ContactGAN takes a predicted protein contact map as input and outputs a refined, more accurate contact map. On a test set of 43 protein domains from CASP13, ContactGAN showed an average improvement of 24% in precision values of L/1 long contacts.

Speakers:

Daisuke Kihara

Purdue University

11:20 AM - 11:40 AM (EDT)

NetBio: Network Biology - Chromatin network markers of leukemia

Motivation: The structure of chromatin impacts gene expression. Its alteration has been shown to coincide with the occurrence of cancer. A key challenge is in understanding the role of chromatin structure in cellular processes and its implications in diseases. Results: We propose a comparative pipeline to analyze chromatin structures and apply it to study chronic lymphocytic leukemia (CLL). We model the chromatin of the affected and control cells as networks and analyze the network topology by state-of-the-art methods. Our results show that chromatin structures are a rich source of new biological and functional information about DNA elements and cells that can complement protein-protein and co-expression data. Importantly, we show the existence of structural markers of cancer-related DNA elements in the chromatin. Surprisingly, CLL driver genes are characterized by specific local wiring patterns not only in the chromatin structure network of CLL cells, but also of healthy cells. This allows us to successfully predict new CLL-related DNA elements. Importantly, this shows that we can identify cancer-related DNA elements in other cancer types by investigating the chromatin structure network of the healthy cell of origin, a key new insight paving the road to new therapeutic strategies. This gives us an opportunity to exploit chromosome conformation data in healthy cells to predict new drivers.

Speakers:

Noel Malod-Dognin

Barcelona Supercomputing Center (BSC)

11:20 AM - 11:30 AM (EDT)

CompMS - CoronaMassKB: an open-access platform for sharing of mass spectrometry data and reanalyses from SARS-CoV-2 and related species

CoronaMassKB is an open-data community resource for sharing of mass spectrometry data and (re)analysis results for all experiments pertinent to the global SARS-CoV-2 pandemic. The resource is designed for the rapid exchange of data and results among the global community of scientists working towards understanding the biology of SARS-CoV-2/COVID19 and thus aims to accelerate the emergence of effective responses to this global pandemic. CoronaMassKB is currently based on reanalysis of >7 million spectra from SARS-CoV-2 dataset and >20 million spectra from related viruses (including SARS-Cov, MERS and H1N1) in public mass spectrometry data available at MassIVE. Searching for modified peptides nearly doubled the number of identifications derived from the same data and reveals >550 SARS-CoV-2 hypermodified peptides with 10+ distinct combinations of modifications, up to over 100 unique modification variants for a single peptide sequence. Overall, thousands of distinctly-modified and unmodified peptide variants were identified to 25 SARS-CoV-2 protein products, with thousands of additional variants also identified for the various related viral species. Over 7,000 host proteins were also detected by tens of thousands of peptide variants across several types of experiments, including 339 variants mapped to regulatory regions of 92 drug-target proteins assessed to be interactors of SARS-CoV-2 proteins.

Speakers:

Nuno Bandeira

University of California San Diego

11:20 AM - 11:40 AM (EDT)

Vari - Population-specific VNTR Alleles in the Human Genome

Variable number of tandem repeats (VNTRs) are polymorphic DNA tandem repeat loci in which the number of pattern copies varies across a population. Human minisatellite VNTR loci (with pattern lengths from seven to hundreds of base pairs) have a variety of functional effects (transcription factor binding, RNA splicing) and are associated with disease (neurodegenerative disorders, cancers, Alzheimer’s disease). Despite their importance, relatively few minisatellite VNTRs have been identified and studied in detail. As part of a large survey of VNTR occurrence in over 2,500 human whole genome sequencing samples from the 1000 genomes project, we sought to identify population-specific VNTR alleles. We found 5,541 “common” VNTR loci (occurring in ≥ 5% of the samples) and used their alleles to develop a decision tree classification model to predict super-population membership, with 97.81% accuracy. We then identified 1,283 top population predictive alleles. Finally, we developed a novel ‘Virtual Gel’ illustration showing how alleles differ across populations at population-specific loci. This is the first large-scale study of population-specific VNTRs and the information obtained could be useful for haplotype inference, studies of human migration and evolution, and accurate use of VNTRs in GWAS studies.

Speakers:

Marzieh Eslami Rasekh

Boston University

11:20 AM - 11:40 AM (EDT)

TransMed - Privacy-preserving Construction of Generalized Linear Mixed Model for Biomedical Computation

The Generalized Linear Mixed Model (GLMM) is an extension of the generalized linear model (GLM) in which the linear predictor takes into account random effects.Given its power of precisely modeling the mixed effects from multiple sources of random variations, the method has been widely used in biomedical computation, for instance in the genome-wide association studies (GWAS) that aim to detect genetic variance significantly associated with phenotypes such as human diseases. Collaborative GWAS on large cohorts of patients across multiple institutions is often impeded by the privacy concerns of sharing personal genomic and other health data. To address such concerns, we present in this paper a privacy-preserving Expectation-Maximization (EM) algorithm to build GLMM collaboratively when input data are distributed to multiple participating parties and cannot be transferred to a central server. We assume that the data are horizontally partitioned among participating parties: i.e., each party holds a subset of records (including observational values of fixed effect variables and their corresponding outcome), and for all records, the outcome is regulated by the same set of known fixed effects and random effects. Our collaborative EM algorithm is mathematically equivalent to the original EM algorithm commonly used in GLMM construction. The algorithm also runs efficiently when tested on simulated and real human genomic data, and thus can be practically used for privacy-preserving GLMM construction.

Speakers:

Rui Zhu

Indiana University

11:20 AM - 11:40 AM (EDT)

RegSys - Hox binding specificity is directed by DNA sequence preferences and differential abilities to engage inaccessible chromatin

The Hox transcription factors (TFs) bind similar target sequence motifs in vitro, yet they specify distinct cellular fates in the developing embryo. How the Hox TFs achieve the specificity required in order to drive differential cell fates remains unclear. Here, we induce the expression of several Hox TFs in the uniform cellular context of embryonic stem cell derived progenitor motor neurons (pMNs). We ask whether Hox TFs bind differentially in pMNs, and whether differential Hox TF binding is determined by sequence preferences or the ability of Hox TFs to differentially interact with preexisting pMN chromatin. We find that Hox TFs bind both “shared” and “unique” sites in the genome, and sequence preferences are unable to explain the observed diversity in Hox TF binding. To examine the degree to which preexisting chromatin predicts induced Hox binding, we develop a novel bimodal neural network framework that estimates the degree to which sequence and pMN chromatin features predict induced Hox TF binding. Results from our network suggest that the preexisting chromatin environment determines the binding of different Hox TFs to varying degrees. For Hox TFs that bind similar sequences, differential abilities to bind inaccessible chromatin drive differential Hox TF binding.

Speakers:

Divyanshi Srivastava

Center for Eukaryotic Gene Regulation, The Pennsylvania State University, University Park, PA

11:20 AM - 11:30 AM (EDT)

Bio-Ontologies - CORAL: A platform for FAIR, rigorous, self-validated data modeling and integrative, reproducible data analysis

Many organizations face challenges in managing and analyzing data, especially when such data is obtained from multiple sources, created using diverse methods or protocols. Analyzing heterogeneous, structured datasets requires rigorous tracking of their interrelationships and provenance. This task has long been a Grand Challenge of data science, and has more recently been formalized in the FAIR principles: that all data be Findable, Accessible, Interoperable and Reusable, both for machines and for people. Adherence to these principles is necessary for proper stewardship of information, for testing regulatory compliance, for measuring efficiency, and for effectively being able to reuse data analytical frameworks. Interoperability and reusability are especially challenging to implement in practice, to the extent that scientists acknowledge a “reproducibility crisis” across many fields of study. We developed CORAL, a framework for organizing the large diversity of datasets that are generated and used by moderately complex organizations. CORAL features a web interface for bench scientists to upload and explore data, as well as a Jupyter notebook interface for data analysts, both backed by a common API. We describe the CORAL data model and associated tools, and discuss how they greatly facilitate adherence to all four of the FAIR principles.

Speakers:

John-Marc Chandonia

Berkeley National Lab

11:20 AM - 11:30 AM (EDT)

3DSIG - Deep Learning Protein Contacts and Real-valued Distances Using PDNET

As deep learning algorithms drive the progress in protein structure prediction, a lot remains to be studied at this emerging crossway of deep learning and protein structure prediction. Recent findings show that inter-residue distance prediction, a more granular version of the well-known contact prediction problem, is a key to predict accurate models. We believe that deep learning methods that predict these distances are still at infancy. To advance these methods and develop other novel methods, we need a small and representative dataset packaged for fast development and testing. In this work, we introduce Protein Distance Net (PDNET), a dataset derived from the widely used DeepCov dataset consisting of 3456 representative protein chains. It is packaged with all the scripts that were used to curate the dataset, generate the input features and distance maps, and scripts with deep learning models to train, validate, and test. Deep learning models can also be trained and tested in a web browser using free platforms such as Google Colab. We discuss how our framework can be used to predict contacts, distance intervals, and real-valued distances. PDNET is available at https://github.com/ba-lab/pdnet/.

Speakers:

Badri Adhikari

12:00 PM - 12:40 PM (EDT)

3DSIG - 3DSIG Keynote: Thinking Deeply About Protein Structure Prediction

In this talk I will overview the astonishing recent progress in protein structure prediction that has arisen from better modelling of amino acid covariation effects, and most recently, the application of deep neural networks to the problem e.g. with Deepmind's AlphaFold. Bringing things up to the present day, I will discuss some of the recent method developments we have been experimenting with in my own lab, and discuss how close we are to being able to model the structure for the entire complement of proteins encoded by a bacterial genome, and the interactions between them. I will finish by discussing the future prospects for these techniques and discuss some of the current limitations, along with some intriguing new methods that are coming down the line that might be able to take us further.

Speakers:

David Jones

University College London

12:00 PM - 12:20 PM (EDT)

NetBio: Network Biology - Prediction of cancer driver genes through network-based moment propagation of mutation scores

Motivation: Gaining a comprehensive understanding of the genetics underlying cancer development and progression is a central goal of biomedical research. Its accomplishment promises key mechanistic, diagnostic and therapeutic insights. One major step in its direction is the identification of genes that drive the emergence of tumors upon mutation. Recent advances in the field of computational biology have shown the potential of combining genetic summary statistics that represent the mutational burden in genes with biological networks, such as protein-protein interaction networks, to identify cancer driver genes. Those approaches superimpose the summary statistics on the nodes in the network, followed by an unsupervised propagation of the node scores through the network. However, this unsupervised setting does not leverage any knowledge on well-established cancer genes, a potentially valuable resource to improve the identification of novel cancer drivers. Results: We develop a novel node embedding that enables classification of cancer driver genes in a supervised setting. The embedding combines a representation of the mutation score distribution in a node’s local neighborhood with network propagtion. We leverage the knowledge of well-established cancer driver genes to define a positive class, resulting in a partially-labeled data set, and develop a cross validation scheme to enable supervised prediction. The proposed node embedding followed by a supervised classification improves the predictive performance compared to baseline methods, and yields a set of promising genes that constitute candidates for further biological validation.

Speakers:

Anja Gumpinger

ETH Zurich

12:10 PM - 12:20 PM (EDT)

BioVis - Grid-Constrained Dimensionality-Reduction for Single-Cell RNA-Seq Summarization

Single-cell transcriptomics has become an increasingly common technique for understanding complex biological systems. Due to the high-dimensional data that is generated by such experiments, algorithms for dimensionality reduction, such as principal component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), and uniform manifold approximation and project (UMAP), attempt to overcome these problems through projection of the high-dimensional findings into a lower-dimensional space. As the output of these methods are commonly presented as scatterplots, they often suffer from overplotting, making nearby points difficult to distinguish. Since the spacing between points can be non-uniform, estimation of the relative proportions of cells in each cluster is not practical. Lack of consistent boundaries between points can also make it challenging to interpret cluster membership along with additional parameters, such as gene expression, in the same plot. Here we present a complementary method to these clustering techniques for single cell data. By using linear assignment to map these projections to an appropriately-sized grid, it becomes possible to preserve overall cell-cell relationships in space, while condensing the space used for the figure, avoiding overplotting, and allowing for easy boundary annotation that permits overlay of additional data

Speakers:

Jon Hill

Boehringer Ingelheim Pharmaceuticals Inc

12:20 PM - 12:40 PM (EDT)

Bio-Ontologies - CIDO Diagnosis: COVID-19 diagnosis modeling, representation and analysis using the Coronavirus Infectious Disease Ontology

Diagnosis of COVID-19 is critical to the control of COVID-19 pandemic. Common diagnostic methods include symptoms identification, chest imaging, serological test, and RT-PCR. However, the sensitivity and specificity of different diagnosis methods differ. In this study, we ontologically represent different aspects of COVID-19 diagnosis using the community- based Coronavirus Infectious Disease Ontology (CIDO), an OBO Foundry library ontology. CIDO includes many new terms and also imports many relevant terms from existing ontologies. The high level hierarchy and design pattern of CIDO are introduced to support COVID-19 diagnosis. The knowledge reported in the literature reports and reliable resources such as the FDA website is ontologically represented. We modeled and compared over 20 SARS-CoV-2 RT-PCR assays, which target different gene markers in SARS-CoV-2. The sensitivity and specificity of different methods are discussed.

Speakers:

Yongqun He

University of Michigan Medical School

12:20 PM - 12:30 PM (EDT)

TransMed - PRECISE+ predicts drug response in patients by non-linear subspace based transfer from cell lines and PDX models

Pre-clinical models have extensively been used to understand the molecular underpinnings of cancer. Cell lines and Patient Derived Xenografts (PDX) are amenable to screening for a wide range of anti-cancer therapeutics. These screens offer a direct measure of drug response for many drugs – data that cannot be collected for human tumors. Pre-clinical models do, however, show behavioral discrepancies with respect to human tumors that impedes the transfer of biomarkers of drug response from pre-clinical models to patients. We present a novel framework for integrating omics data derived from pre-clinical models and tumors. Our approach employs non-linear dimensionality reduction to capture complex genetic interaction patterns that are common to pre-clinical models and humans. These patterns are then used to train a drug response predictor on human tumor data. This work extends PRECISE to allow incorporation of non-linear similarity measures between samples while retaining equivalence to the linear setting.

Speakers:

Soufiane Mourragui

Delft University of Technology and the Netherlands Cancer Institute

12:20 PM - 12:40 PM (EDT)

NetBio: Network Biology - Network-principled deep generative models for designing drug combinations as graph sets

Motivation: Combination therapy has shown to improve therapeutic efficacy while reducing side effects. Importantly, it has become a indispensable strategy to overcome resistance in antibiotics, anti-microbials, and anti-cancer drugs. Facing enormous chemical space and unclear design principles for small-molecule combinations, computational drug-combination design has not seen generative models to meet its potential to accelerate resistance-overcoming drug combination discovery. Results: We have developed the first deep generative model for drug combination design, by jointly embedding graph-structured domain knowledge and iteratively learning a reinforcement learning-based chemical graph-set designer. First, we have developed Hierarchical Variational Graph Auto-Encoders (HVGAE) trained end-to-end to jointly embed gene-gene, disease-disease and gene-disease networks. Novel attentional pooling is introduced here for learning disease-representations from associated genes' representations. Second, targeting diseases in learned representations, we have recast the drug-combination design problem as graph-set generation and developed a deep learning-based model with novel rewards. Specifically, besides chemical validity rewards, we have introduced novel generative adversarial award, being generalized sliced Wasserstein, for chemically diverse but distributionally similar molecules to known drug-like compounds or drugs. We have also designed a network principle-based reward for drug combinations. Numerical results indicate that, compared to state-of-the-art graph embedding methods, HVGAE learns more generalizable and informative disease representations in disease-disease graph reconstruction. Results also show that the deep generative models generate drug combinations following the principle across diseases. A case study on melanoma shows that generated drug combinations collectively cover the disease module similar to FDA-approved drug combinations and could also suggest promising novel systems-pharmacology strategies. Our method allows for examining and following network-based principle or hypothesis to efficiently generate disease-specific drug combinations in a vast chemical combinatorial space.

Speakers:

Mostafa Karimi

Texas A&M University

12:20 PM - 12:40 PM (EDT)

RegSys - Fully Interpretable Deep Learning Model of Transcriptional Control

The universal expressibility assumption of Deep Neural Networks (DNNs) is the key motivation behind recent works in the system biology community to employ DNNs to solve important problems in functional genomics and molecular genetics. Typically, such investigations have taken a "black box" approach in which the internal structure of the model used is set purely by machine learning considerations with little consideration of representing the internal structure of the biological system by the mathematical structure of the DNN. DNNs have not yet been applied to the detailed modeling of transcriptional control in which mRNA production is controlled by the binding of specific transcription factors to DNA, in part because such models are in part formulated in terms of specific chemical equations that appear different in form from those used in neural networks. In this paper, we give an example of a DNN which can model the detailed control of transcription in a precise and predictive manner. Its internal structure is fully interpretable and is faithful to underlying chemistry of transcription factor binding to DNA. We derive our DNN from a systems biology model that was not previously recognized as having a DNN structure. Although we apply our DNN to data from the early embryo of the fruit fly Drosophila, this system serves as a testbed for analysis of much larger data sets obtained by systems biology studies on a genomic scale.

Speakers:

Yi Liu

University of Chicago

12:20 PM - 12:40 PM (EDT)

HiTSeq: High Throughput Sequencing - Weighted minimizer sampling improves long read mapping

In this era of exponential data growth, minimizer sampling has become a standard algorithmic technique for rapid genome sequence comparison. This technique yields a sub-linear representation of sequences, enabling their comparison in reduced space and time. A key property of the minimizer technique is that if two sequences share a substring of a specified length, then they can be guaranteed to have a matching minimizer. However, because the k-mer distribution in eukaryotic genomes is highly uneven, minimizer-based tools (e.g., Minimap2, Mashmap) opt to discard the most frequently occurring minimizers from the genome in order to avoid excessive false positives. By doing so, the underlying guarantee is lost and accuracy is reduced in repetitive genomic regions. We introduce a novel weighted-minimizer sampling algorithm. A unique feature of the proposed algorithm is that it performs minimizer sampling while taking into account a weight for each k-mer; i.e, the higher the weight of a k-mer, the more likely it is to be selected. By down-weighting frequently occurring k-mers, we are able to meet both objectives: (i) avoid excessive false-positive matches, and (ii) maintain the minimizer match guarantee. We tested our algorithm, Winnowmap, using both simulated and real long-read data and compared it to a state-of-the-art long read mapper, Minimap2. Our results demonstrate a reduction in the mapping error-rate from 0.14% to 0.06% in the recently finished human X chromosome (154.3 Mbp), and from 3.6% to 0% within the highly repetitive X centromere (3.1 Mbp). Winnowmap improves mapping accuracy within repeats and achieves these results with sparser sampling, leading to better index compression and competitive runtimes.

Speakers:

Chirag Jain

12:30 PM - 12:40 PM (EDT)

TransMed: Q&A

Speakers:

Center for Algorithmic Biotechnology, Institute of Translational Biomedicine, St. Petersburg State University

01:00 PM - 01:30 PM (EDT)

Lunch & Learn - The Black Women in Computational Biology Network

Speakers:

Jenea Adams

University of Pennsylvania

02:00 PM - 03:00 PM (EDT)

BioVis Keynote: Machine Learning for Drug Repurposing

Speakers:

Marinka Zitnik

Harvard

02:00 PM - 02:10 PM (EDT)

3DSIG - De Novo Protein Design for Novel Folds with Guided & Conditional Wasserstein GAN

Facing data quickly accumulating on protein sequence and structure, this study is addressing the following question: to what extent could current data alone reveal deep insights into the sequence-structure relationship, such that new sequences can be designed accordingly for novel structure folds? We have developed novel deep generative models, constructed low-dimensional representation of fold space, exploited sequence data with and without paired structures, and developed ultra-fast fold predictor (oracle) as feedback. The resulting semi-supervised gcWGAN (guided & conditional, Wasserstein Generative Adversarial Networks), assessed by the oracle over 100 novel folds not in the training set, generates more yields and covers 3.5 times more target folds compared to a competing data-driven method (cVAE). gcWGAN designs are predicted to be physically and biologically sound. Targeting representative novel folds, including one not even part of basis folds, gcWGAN designs are predicted by Rosetta to have comparable or better fold accuracy; yet have much more sequence diversity and sometimes novelty. The ultra-fast data-driven model is shown to boost the success of principle-driven Rosetta for de novo design, through generating design seeds and tailoring design space. In conclusion, gcWGAN explores uncharted sequence space to design proteins by learning from current sequence-structure data.

Speakers:

Yang Shen

Texas A&M University

02:00 PM - 02:20 PM (EDT)

HiTSeq: High Throughput Sequencing - PopDel detects large deletions jointly in tens of thousands of genomes

Catalogs of genetic variation for large numbers of individuals are a foundation for modern research on human diversity and disease. Creating such catalogs for small variants from whole-genome sequencing (WGS) data is now commonly done for thousands of individuals collectively. We have transferred this joint calling idea from SNPs and indels to larger deletions and developed the first joint calling tool, PopDel, that can detect and genotype deletions in WGS data of tens of thousands of individuals simultaneously as demonstrated by our evaluation on data of up to 49,962 human genomes. Good sensitivity, precision and the correctness of genotypes are demonstrated by extensive tests on simulated and real data and comparison to other state-of-the-art SV-callers. PopDel detects deletions in HG002 and NA12878 with high sensitivity while maintaining a low false positive rate as shown by our comparison with different high-confidence reference sets. On data of up to 6,794 trios, inheritance patterns are in concordance with Mendelian inheritance rules and exhibit a close to ideal transmission rate. PopDel reliably reports common, rare and de novo deletions. Therefore, PopDel enables routine scans for deletions in large-scale sequencing studies and we are currently in the process of implementing the detection of other SV-types.

Speakers:

Sebastian Niehus

Berlin Institute of Health (BIH) / Charité – Universitätsmedizin Berlin

02:00 PM - 02:20 PM (EDT)

Bio-Ontologies - Modeling quantitative traits for COVID-19 case reports

Medical practitioners record the condition status of a patient through qualitative and quantitative observations. The measurement of vital signs and molecular parameters in the clinics gives a complementary description of abnormal phenotypes associated with the progression of a disease. The Clinical Measurement Ontology (CMO) is used to standardize annotations of these measurable traits. However, researchers have no way to describe how these quantitative traits relate to phenotype concepts in a machine-readable manner. Using the WHO clinical case report form standard for the COVID-19 pandemic, we modeled quantitative traits and developed OWL axioms to formally relate clinical measurement terms with anatomical, biomolecular entities and phenotypes annotated with the Uber-anatomy ontology (Uberon), Chemical Entities of Biological Interest (ChEBI) and the Phenotype and Trait Ontology (PATO) biomedical ontologies. The formal description of these relations allows interoperability between clinical and biological descriptions, and facilitates automated reasoning for analysis of patterns over quantitative and qualitative biomedical observations.

Speakers:

Núria Queralt-Rosinach

Leiden University Medical Center

Ontario Institute for Cancer Research

02:10 PM - 02:20 PM (EDT)

3DSIG - The vestibule role of membrane-water interface as the intermediate stage in a new three-stage model for helical membrane protein folding

Transmembrane alpha-helical (TMH) proteins play critical roles in cellular signaling. They display a diversity of structural folds featuring almost-parallel orientation of TM helices packing into helical bundles. The membrane environment enormously reduces the accessible conformational landscape for folding, but also makes its experiments challenging. The contribution of helix insertion energies to the folding energy landscape was computed using structural bioinformatics based hydropathy analysis for most of the polytopic helical membrane proteome (from 1-TMH to 24-TMH proteins with structures). The magnitudes of TM helix insertion energies from Water to membrane-water Interface (WAT→INT energies) are on average half of those insertion energies from water to Trans-Membrane-Helix orientation (WAT→TMH energies), suggesting a potential vestibule role of the membrane-water interface for the TM helices after translocon exit. This is confirmed by showing the stability of very hydrophobic TM helices in the membrane-water interface through multiple microsecond long molecular dynamics simulations of a stop-transfer helix, a re-integration helix, and a pre-folded helical-hairpin from the ribosomal exit vestibule. So, a three-stage folding model is proposed to extend Popot-Engelman’s original two-stage model, where the membrane-water interface acts as the intermediate stage holding vestibule for translated TM helices, reconciling the interface’s critical role seen in many previous studies.

Speakers:

Ravinder Abrol

California State University, Northridge

02:20 PM - 02:40 PM (EDT)

3DSIG - Detecting Symmetry in Membrane Proteins

Available membrane protein structures have revealed an abundance of symmetry and pseudo-symmetry, which arose not only by the formation of multi-subunit assemblies, but also by repetition of internal structural elements. In many cases, these symmetry relationships play a crucial role in defining the functional properties of the proteins. Therefore, a systematic study of symmetry should provide a framework for a broader understanding of the mechanistic principles and evolutionary development of membrane proteins. However, available symmetry detection methods have not been tested systematically on this class of proteins because of the lack of an appropriate benchmark set. Hence, we collected membrane protein structures with unique architectures and manually curated their symmetries to create the MemSTATS dataset. Using MemSTATS, we compared the performance of four widely used symmetry detection algorithms and pinpointed areas for improvement. To address the identified shortcomings, we developed a robust symmetry detection methodology called MSSD, which takes into consideration the restrictions that the lipid bilayer places on protein structures. MSSD detected symmetries with higher accuracy and lower false positive rate compared to any other tested method. Consequently, we used MSSD to analyze all available membrane protein structures and presented the resultant symmetries in a database called EncoMPASS (encompass.ninds.nih.gov).

Speakers:

Antoniya Aleksandrova

NIH - NINDS

02:20 PM - 02:30 PM (EDT)

TransMed - Deep Hidden Physics Modeling of Cell Signaling Networks

Signaling systems in multicellular organisms are vital for cell-cell communication, tissue organization and disease. Cancer genomics has unraveled a surprisingly large set of novel gene lesions from tumors. Our previous studies have globally explored the rewiring of cell signaling networks underlying malignant transformation caused by kinases and other signaling proteins. By generating quantitative time-/state-series data and subsequently using these as input for deep learning based computational modeling - our lab work to identify the principal changes in the genome, cell signaling and phenotypes of cells harboring genetic mutations; we validate these models by forward prediction of experimentally observed phenotypic responses to drug and genetic perturbations. We are currently deploying such forecasting models on data collected from PDX tumors to describe how the cell signaling networks are mechanistically, dynamically and differentially utilized in cancers. Finally, we are working to combine deep learning with causal/mechanistic models to predict novel treatment and diagnostic strategies for tumors harboring different genetic lesions. In conclusion, our studies aim to unravel the fundamental rewiring of cell signaling networks in cancer and will serve as a major breakthrough in our basic understanding of their impact on the disease, paving the way for future clinical applications and tumor specific cancer therapy.

Speakers:

Rune Linding

Humboldt-Universität zu Berlin

02:20 PM - 02:40 PM (EDT)

HiTSeq: High Throughput Sequencing - Metalign: Efficient alignment-based metagenomic profiling via containment min hash

Whole-genome shotgun sequencing enables the analysis of microbial communities in unprecedented detail, with important implications in medicine and ecology. Predicting the presence and relative abundances of microbes in a sample, known as “metagenomic profiling”, is a critical step in microbiome analysis. Existing profiling methods have been shown to suffer from poor false positive or false negative rates, while alignment-based approaches are often considered accurate but computationally infeasible. Here we present a novel method, Metalign, that addresses these concerns by performing efficient alignment-based metagenomic profiling. Metalign employs a high-speed, high-recall pre-filtering method based on the mathematical concept of Containment Min Hash to reduce the reference database size dramatically before alignment, followed by a method to estimate organism relative abundances in the sample by handling reads aligned to multiple genomes. We show that Metalign achieves significantly improved results over existing methods on simulated datasets (Figure 1) from a large benchmarking study, CAMI, and performs well on in vitro mock community data and environmental data from the Tara Oceans project. Metalign is freely available at https://github.com/nlapier2/Metalign, and via bioconda.

Speakers:

Nathan Lapierre

University of California, Los Angeles

02:20 PM - 02:40 PM (EDT)

Bio-Ontologies - Using ontologies to extract disease-phenotype associations from literature

With the advances in Next Generation Sequencing (NGS) technologies, a huge volume of clinical genomic data has become available. Efficient exploitation of such data requires linkage to a patient's complete phenotype profile. Current resources providing disease-phenotype associations are not comprehensive, and they often do not cover all of the diseases from OMIM and particularly from ICD10, which are the primary terminologies used in clinical settings. Here, we propose a text-mining system which utilizes semantic relations in the phenotype ontologies and statistical methods to extract disease-phenotype associations from the literature. We compare our findings against established disease-phenotype associations and also demonstrate its utility in covering mouse gene-disease associations from Mouse Genome Informatics (MGI). Such associations serve as necessary information blocks for understanding underlying disease mechanisms and developing or repurposing drugs.

Speakers:

Senay Kafkas

King Abdullah University Of Science and Technology

02:30 PM - 02:40 PM (EDT)

TransMed - A deep transfer learning model for extending in vitro CRISPR-Cas9 viability screens to tumors

The Cancer Dependency Map (DepMap) projects recently employed genome-scale CRISPR-Cas9 loss-of-function screens to identify genes essential for cancer cell proliferation and survival across cancer cell lines. However, it remains very challenging to translate these in vitro results to impracticable-to-screen tumors. To address the challenge, we devised a deep learning model with a unique transfer-learning framework to predict gene dependencies of tumors. The model has a 3-stage design that enables a representation learning of unlabeled tumor genomic data, the prediction of gene dependencies in labeled cell-line screening data, and the application to predict tumor dependencies. The prediction performance was verified using cell-line data. Applying our model to ~8,000 tumors of The Cancer Genome Atlas, we constructed a pan-cancer dependency map of tumors. The results were confirmed by several biomarkers and the response to targeted therapies of the TCGA clinical records. Further investigations revealed gene dependencies associated with specific genomic patterns, such as higher tumor mutation burdens and unique expression/methylation signatures. We also identified highly selective gene dependencies of which inhibitor drugs have been approved to treat cancers. We expect the model to evolve with rapidly developing in vitro CRISPR-Cas9 viability screens and facilitate the translation to identifying therapeutic targets of tumors.

Speakers:

Yu-Chiao Chiu

02:40 PM - 03:00 PM (EDT)

HiTSeq: High Throughput Sequencing - META^2: Memory-efficient taxonomic classification and abundance estimation for metagenomics with deep learning

Taxonomic classification is an important step in the analysis of samples found in metagenomic studies. Conventional mapping-based methods trade off between high memory and low recall, with recent deep learning methods suffering from very large model sizes. We aim to develop a more memory-efficient technique for taxonomic classification. A task of particular interest is abundance estimation. Current methods initially classify reads independently and are agnostic to the co-occurence patterns between taxa. In this work, we also attempt to take these patterns into account. We develop a novel memory-efficient read classification technique, combining deep learning and locality-sensitive hashing. We show that this approach outperforms conventional mapping-based and other deep learning methods for taxonomic classification when restricting all methods to a fixed memory footprint. Moreover, we formulate the task of abundance estimation as a Multiple Instance Learning problem and we extend current deep learning architectures with two types of permutation-invariant MIL pooling layers: a) deepsets and b) attention-based pooling. We illustrate that our architectures can exploit the co-occurrence of species in metagenomic read sets and outperform the single-read architectures in predicting the distribution over taxa at higher taxonomic ranks.

Speakers:

Harun Mustafa

ETH Zürich

02:50 PM - 03:00 PM (EDT)

TransMed: Q&A

Speakers:

Macquarie University

03:20 PM - 03:40 PM (EDT)

CompMS - Isolation forests improve the capability to detect quality problems in mass spectrometry-based proteomics

Quality control (QC) of mass spectrometry based proteomic experiments involve quantifying a standard mixture which includes a set of analytes, generating multiple metrics. The metrics are then used to evaluate effects of technical variability to the quantification of the actual biological samples. Next, a reliable baseline data set is used to train statistical models or traditional statistical quality control methods to classify outlying runs. Although current technological improvements help conduct initial steps of most QC workflows, many practitioners still lack baseline data and misclassifies QC runs in real time implementation. Here, we present an unsupervised machine learning extension of MSstatsQC to detect deviations from optimal performance of multiple metrics. MSstatsQC implements unsupervised isolation-based trees to address outlier detection problem. Our results show how tree-based methods are helpful in terms of differentiating optimal and sub-optimal experiments where limited information is available about the optimal/suboptimal performance of the instrument. We also provide supporting information based on the root causes of anomalous behavior per peptide which can be used to design preventive actions. Our method is available with MSstatsQC R/Bioconductor package and with web-based graphical user interface MSstatsQCgui.

Speakers:

Eralp Dogu

College of Science, Mugla Sitki Kocman University

03:20 PM - 03:40 PM (EDT)

NetBio: Network Biology - Understanding tissue-specific gene regulation by miRNAs

Conventional methods to analyze genomic data do not make use of the interplay between multiple factors, such as between microRNAs (miRNAs) and the mRNA transcripts they regulate, and thereby often fail to identify the cellular processes that are unique to specific tissues. We developed PUMA (PANDA Using MicroRNA Associations), a computational tool that uses message passing to integrate a prior network of miRNA target predictions with protein-protein interaction and target gene co-expression information to model genome-wide gene regulation by miRNAs. We applied PUMA to 38 tissues from the Genotype-Tissue Expression project, integrating RNA-Seq data with two different miRNA target predictions priors, built on predictions from TargetScan and miRanda, respectively. We found that while target predictions obtained from these two different resources are considerably different, PUMA captures similar tissue-specific miRNA-target gene regulatory interactions in the different network models. Furthermore, tissue-specific functions of miRNAs, which we identified by analyzing their regulatory profiles, are highly similar between networks modeled on the two target prediction resources. This indicates that PUMA consistently captures important tissue-specific regulatory processes of miRNAs. In addition, using PUMA we identified miRNAs regulating important tissue-specific processes that, when mutated, may result in disease development in the same tissue.

Speakers:

Marieke Kuijjer

Centre for Molecular Medicine Norway (NCMM), University of Oslo

03:20 PM - 03:40 PM (EDT)

Bio-Ontologies - Metadata standards for the FAIR sharing of vector embeddings in Biomedicine

Motivation: Today, we have an enormous amount of biomedical data and its size, as well as complexity, have been increasing over time. Implementation of standards represents one of the key drivers in the life sciences research as well as the technology transfer. More specifically, standards enable data accessibility, sharing, integration and therefore facilitates data harnessing and accelerates research and innovation transfer. The life sciences community has widely developed and used Semantic web technology standards for data representation and sharing. However, given the success of unsupervised machine learning methods such as Word2Vec and BERT, there is a need to develop new standards for sharing the (pre-trained) vector space embeddings of the entities to facilitate reusability of data and method development. Motivated by this, we propose data and metadata standards for the FAIR distribution of vector embeddings and demonstrate utilization of these standards in Bio2Vec, a platform providing a flexible, reliable and standard-compliant data representation, sharing, integration and analysis. Availability: The proposed metadata standard and an example are available in the ShEx format at Zenodo.

Speakers:

Remzi Çelebi

Maastricht University

03:20 PM - 03:40 PM (EDT)

TransMed - Identifying diagnosis-specific genotype-phenotype associations via joint multi-task sparse canonical correlation analysis and classification

Brain imaging genetics provides us a new opportunity to understand the pathophysiology of brain disorders. It studies the complex association between genotypic data such as single nucleotide polymorphisms (SNPs) and imaging quantitative traits (QTs). The neurodegenerative disorders usually exhibit the diversity and heterogeneity, originating from which different diagnostic groups might carry distinct imaging QTs, SNPs and their interactions. Sparse canonical correlation analysis (SCCA) is widely used to identify bi-multivariate genotype-phenotype associations. However, most existing SCCA methods are unsupervised, leading to an inability to identify diagnosis-specific genotype-phenotype associations. In this paper, we propose a new joint multi-task learning method, named MT-SCCALR, which absorbs the merits of both SCCA and logistic regression. MT-SCCALR learns genotype-phenotype associations of multiple tasks jointly, with each task focusing on identifying the task-specific genotype-phenotype pattern. To ensure the interpretation and stability, we endow the proposed model with the selection of SNPs and imaging QTs for each diagnostic group alone, while allowing the selection of them shared by multiple diagnostic groups. We derive an efficient optimization algorithm whose convergence to a local optimum is guaranteed. Compared with two state-of-the-art methods, the results show that MT-SCCALR yields better or similar canonical correlation coefficients (CCCs) and classification performances. In addition, it owns much better discriminative canonical weight patterns of great interest than competitors. This demonstrates the power and capability of MTSCCAR in identifying diagnostically heterogeneous genotype-phenotype patterns, which would be helpful to understand the pathophysiology of brain disorders.

Speakers:

Lei Du

Northwestern Polytechnical University

03:20 PM - 03:40 PM (EDT)

MICROBIOME - ganon: precise metagenomics classification against large and up-to-date sets of reference sequences

The exponential growth of assembled genome sequences greatly benefits metagenomics studies. However, currently available methods struggle to manage the increasing amount of sequences and their frequent updates. Indexing the current RefSeq can take days and hundreds of GB of memory on large servers. Few methods address these issues thus far, and even though many can theoretically handle large amounts of references, time/memory requirements are prohibitive in practice. As a result, many studies that require sequence classification use often outdated and almost never truly up-to-date indices. Motivated by those limitations we created ganon, a k-mer based read classification tool that uses Interleaved Bloom Filters in conjunction with a taxonomic clustering and a k-mer counting/filtering scheme. Ganon provides an efficient method for indexing references, keeping them updated. It requires less than 55 minutes to index the complete RefSeq of bacteria, archaea, fungi and viruses. The tool can further keep these indices up-to-date in a fraction of the time necessary to create them. Ganon makes it possible to query against very large reference sets and therefore it classifies significantly more reads and identifies more species than similar methods. When classifying a high-complexity CAMI challenge dataset against complete genomes from RefSeq, ganon shows strongly increased precision with equal or better sensitivity compared with state-of-the-art tools. With the same dataset against the complete RefSeq, ganon improved the F1-Score by 65% at the genus level. It supports taxonomy- and assembly-level classification, multiple indices and hierarchical classification. The software is open-source and available at: https://gitlab.com/rki_bioinformatics/ganon

Speakers:

Vitor Piro

Hasso Platner Institute

03:20 PM - 03:40 PM (EDT)

BioVis - Interactive visualization and analysis of morphological skeletons of brain vasculature networks with VessMorphoVis

Motivation: Accurate morphological models of brain vasculature are key to modeling and simulating cerebral blood flow (CBF) in realistic vascular networks. This in silico approach is fundamental to revealing the principles of neurovascular coupling (NVC). Validating those vascular morphologies entails performing certain visual analysis tasks that cannot be accomplished with generic visualization frameworks. This limitation has a substantial impact on the accuracy of the vascular models employed in the simulation. Results: We present VessMorphoVis, an integrated suite of toolboxes for interactive visualization and analysis of vast brain vascular networks represented by morphological graphs segmented originally from imaging or microscopy stacks. Our workflow leverages the outstanding potentials of Blender, aiming to establish an integrated, extensible and domain-specific framework capable of interactive visualization, analysis, repair, high-fidelity meshing and high-quality rendering of vascular morphologies. Based on the initial feedback of the users, we anticipate that our framework will be an essential component in vascular modeling and simulation in the future, filling a gap that is at present largely unfulfilled. Availability and implementation: VessMorphoVis is freely available under the GNU public license on Github at https://github.com/BlueBrain/VessMorphoVis. The morphology analysis, visualization, meshing and rendering modules are implemented as an add-on for Blender 2.8 based on its Python API (Application Programming Interface). The add-on functionality is made available to users through an intuitive graphical user interface (GUI), as well as through exhaustive configuration files calling the API via a feature-rich CLI (command line interface) running Blender in background mode.

Speakers:

Marwan Abdellah

Blue Brain Project / EPFL

03:20 PM - 03:40 PM (EDT)

Vari - Combinatorial and statistical prediction of gene expression from haplotype sequence

Motivation: Genome-wide association studies have discovered thousands of significant genetic effects on disease phenotypes. By considering gene expression as the intermediary between genotype and disease phenotype, eQTL studies have interpreted these variants by their regulatory effects on gene expression. However, there remains a considerable gap between genotype-to-gene expression association and genotype-to-gene expression prediction. Accurate prediction of gene expression enables gene-based association studies to be performed post-hoc for existing GWAS, reduces multiple testing burden, and can prioritize genes for subsequent experiments. Results: In this work, we develop gene expression prediction methods that relax the independence and additivity assumptions between genetic markers. First, we consider gene expression prediction from a conventional regression perspective and develop the HAPLEXR algorithm which combines haplotype clusterings with allelic dosages. Second, we introduce the new gene expression classification problem, which focuses on identifying expression groups rather than continuous measurements; we formalize the selection of an appropriate number of expression groups using the principle of maximum entropy. Third, we develop the HAPLEXD algorithm that incorporates suffix tree based haplotype sharing with spectral clustering to identify expression classes from haplotype sequences. In both models, we penalize model complexity by prioritizing genetic clusters that indicate significant effects on expression. We compare HAPLEXR and HAPLEXD on five GTEx v8 tissues with three state-of-the-art expression prediction methods. HAPLEXD exhibits significantly higher classification accuracy overall and HAPLEXR shows higher prediction accuracy on a significant subset of genes. These results demonstrate the importance of explicitly modelling non-dosage dependent and intragenic epistatic effects when predicting expression.

Speakers:

Berk Alpay

University of Connecticut

03:20 PM - 03:40 PM (EDT)

RegSys - Comparison of chromatin contacts maps from GAM and Hi-C reveals method specific interactions linked with active and inactive chromatin

Gene expression is functionally coupled with 3D genome configuration. Genome Architecture Mapping (GAM) is a ligation-free, genome-wide method that maps chromatin contacts in 3D, based on measuring the frequency of locus co-segregation from an ensemble of ultra-thin nuclear slices of random orientation. To compare GAM and Hi-C, we used our new high-throughput, multiplexed GAM pipeline to produce a deep dataset from mouse embryonic stem cells, and we devised a procedure to extract contacts preferentially detected by either GAM or Hi-C. Strong contacts enriched in the GAM data contain a 2-fold amplification of feature pairs associated with TF binding (including CTCF), histone marks, and enhancers. In contrast, feature pairs enriched in Hi-C-specific contacts are characterized by heterochromatin marks (H3K9me3 and H3K20me3). In general, genomic regions with increased transcriptional activity often form GAM-detected contacts that are underestimated by Hi-C. We are currently investigating whether the differences can be explained by increased contact multiplicity, which could limit ligation-dependent detection. Our findings expand our current understanding of 3D genome folding and highlight the importance of orthogonal approaches.

Speakers:

Christoph Thieme

MDC Berlin

03:30 PM - 03:40 PM (EDT)

3DSIG - Modeling of G protein-coupled receptor structures : Improving the prediction of loop conformations and the usability of models for structure-based drug design

G protein-coupled receptors (GPCRs) form the largest group of potential drug targets and therefore, the knowledge of their three-dimensional structure is important for rational drug design. Homology modeling serves as a common approach for modeling the transmembrane helical cores of GPCRs, however, these models have varying degrees of inaccuracies that result from the quality of template used. We have explored the extent to which inaccuracies inherent in homology models of the transmembrane helical cores of GPCRs can impact loop prediction. We found that loop prediction in GPCR models is much more difficult than loop reconstruction in crystal structures owing to the imprecise positioning of loop anchors. Therefore, minimizing the errors in loop anchors is likely to be critical for optimal GPCR structure prediction. To address this, we have developed a Ligand Directed Modeling (LDM) method comprising of geometric protein sampling and ligand docking. The method was evaluated for capacity to refine the GPCR models built across a range of templates with varying degrees of sequence similarity with the target. LDM reduced the errors in loop anchor positions and improved the prediction of binding poses of ligands, resulting in much better performance of these models in virtual ligand screenings.

Speakers:

Bhumika Arora

Indian Institute of Technology Bombay, Monash University, and IITB-Monash Research Academy

La Jolla Institute of Immunology

03:40 PM - 03:50 PM (EDT)

MICROBIOME - Studying the dynamics of the gut microbiota using metabolically stable isotopic labeling and metaproteomics

The gut microbiome and its metabolic processes are dynamic systems. Surprisingly, our understanding of gut microbiome dynamics is limited. Here we report a metaproteomic workflow that involves protein stable isotope probing (protein-SIP) and identification/quantification of partially labeled peptides. We also developed a package, which we call MetaProfiler, that corrects for false identifications and performs phylogenetic and time series analysis for the study of microbiome dynamics. From the stool sample of five mice that were fed with 15N hydrolysate from Ralstonia eutropha, we identified 15,297 non-redundant unlabeled peptides of which 10,839 of their heavy counterparts were quantified. These results revealed that i) isotope incorporation in proteins differed between taxa, ii) the rate of protein synthesis was lower in the microbiota than in mice, and iii) differences in protein synthesis appeared across protein functions. Interestingly, the phylum Verrucomicrobia and the genera, Akkermansia, Lactobacillus, and Ruminococcus had not reached a plateau of isotopic incorporation 43 days after the continuous introduction of the isotope. Altogether, our study provides an efficient workflow for the study of dynamics of gut microbiota, and our findings helped better understand the complex host-microbiome interactions.

Speakers:

Patrick Smyth

University of Ottawa

03:40 PM - 03:50 PM (EDT)

NetBio: Network Biology - The Reactome Pathway Knowledgebase: Variants, Dark Proteins and Functional Interactions

Reactome is an open access, open source pathway knowledgebase. Its holdings now comprise 12,986 human reactions organized into 2,362 pathways involving 10,908 proteins, 1,865 small molecules, 237 drugs, and 12,206 complexes. 31,237 literature references support these annotations. The roles of variant forms of some proteins, both germline and somatically arising, have been annotated into disease-variant types of reactions and additional reactions that capture the effects of small molecule drugs on these disease processes. To support different visualization and analysis approaches, we implemented several new features through our website, tools, and ReactomeFIViz-Cytoscape app, such as gene set analysis (GSA), an R interface, a Python client, and an intuitive genome-wide results overview based on Voronoi maps. Furthermore, to increase Reactome adoption within the research community, we developed portals and web services for specific user communities. As part of the Illuminating the Druggable (IDG) program, we have undertaken the role to project understudied (Tdark) proteins into the Reactome pathway context, providing useful contextual information for these understudied proteins for experimental biologists to design experiments to understand these proteins’ functions. Reactome thus provides dominant pathway- and network-based tools for analyzing multiple data sets and types.

Speakers:

Robin Haw

03:50 PM - 04:00 PM (EDT)

Vari - A novel framework to evaluate deleteriousness of synonymous variants

Due to the lack of experimental evaluations of variant effects, existing predictors for synonymous singleton nucleotide variants (sSNVs) often rely on databases of variant-disease associations, which are limited for establishing functional effects. Large-scale genome-sequencing efforts have observed roughly four 4 sSNVs (real variants); whereas there are 58 million possible sSNVs have not been observed. We expect that a large fraction of the latter will be found with more sequencing (not-yet-seen variants), while others are either purified due to extreme deleteriousness or unobservable due to physicochemical or sequencing constraints. We further highlight the fact that the not-yet-seen variants are likely similar to the real variants, representing a wide range of functional effects; although the former are probably enriched in more deleterious consequences. We first built a model to identify the not-yet-seen variants from all the unobserved ones. We then trained a second model to differentiate the not-yet-seen and real sets of variants. We further trained the final model model using the real and not-yet-seen variants subset by cutoffs determined by common variants and pathogenic variants, respectively. Our final model outperforms currently available sSNV-predictors in differentiating experimentally verified pathogenic sSNVs from real variants in testing set.

Speakers:

Zishuo Zeng

Rutgers University

04:00 PM - 04:20 PM (EDT)

Bio-Ontologies - hPSCreg-CLO: ontological representation of human pluripotent stem cell lines from the hPSCreg

Human pluripotent stem cells (PSC) are immortal, represent the genotype of the donor and can differentiate into all cell types of a human body. These features establish their enormous potential for modelling diseases and tissues in vitro, drug- and toxicity testing and regenerative medicine. To translate these potencies into reality, large numbers of PS- lines are being generated from a wide spectrum of donors to make them available for the diverse applications. For users to identify suitable PSC- lines, information about the donors, cell generation, characterization and quality are essential. The human pluripotent stem cell registry hPSCreg contains more than 3000 cell line that are richly annotated. To make the hPSCreg resource more accessible and interoperable, we developed hPSCreg-CLO, a new CLO branch that represents various hPSC lines from hPSCreg. hPSC specific design patterns were generated and used to support computer-assisted ontology development. The hPSCreg-CLO includes over 2,400 hPSC lines and their related information such as cell donors, anatomical entities, and original cell types. DL queries were performed to demonstrate the query capability of hPSCreg-CLO. hPSCreg-CLO will further be integrated with the hPSCreg project and support the database data integration and advanced analyses.

Speakers:

Stefanie Seltmann

Charité – Universitätsmedizin Berlin

04:00 PM - 04:20 PM (EDT)

BioVis - ClonArch: Visualizing the Spatial Clonal Architecture of Tumors

Motivation: Cancer is caused by the accumulation of somatic mutations that lead to the formation of distinct populations of cells, called clones. The resulting clonal architecture is the main cause of relapse and resistance to treatment. With decreasing costs in DNA sequencing technology, rich cancer genomics datasets with many spatial sequencing samples are becoming increasingly available, enabling the inference of high-resolution tumor clones and prevalences across different spatial coordinates. While temporal and phylogenetic aspects of tumor evolution, such as clonal evolution over time and clonal response to treatment, are commonly visualized in various clonal evolution diagrams, visual analytics methods that reveal the spatial clonal architecture are missing. Results: This paper introduces ClonArch, a web-based tool to interactively visualize the phylogenetic tree and spatial distribution of clones in a single tumor mass. ClonArch uses the marching squares algorithm to draw closed boundaries representing the presence of clones in a real or simulated tumor. ClonArch enables researchers to examine the spatial clonal architecture of a subset of relevant mutations at different prevalence thresholds and across multiple phylogenetic trees. In addition to simulated tumors with varying number of biopsies, we demonstrate the use of ClonArch on a hepatocellular carcinoma tumor with 280 sequencing biopsies. ClonArch provides an automated way to interactively examine the spatial clonal architecture of a tumor, facilitating clinical and biological interpretations of the spatial aspects of intra-tumor heterogeneity. Availability: https://github.com/elkebir-group/ClonArch

Speakers:

Jiaqi Wu

University of Illinois at Urbana Champaign

04:00 PM - 04:20 PM (EDT)

RegSys - Deciphering the role of 3D genome organization in breast cancer susceptibility

Cancer risk by environmental exposure is modulated by an individual’s genetics and age at exposure. This age-specific period of susceptibility is referred to as a “Window of Susceptibility” (WOS). Radiation exposures poses a high breast cancer risk for women between the early childhood and young adult stage and is reduced in the mid-30s. Rats have a similar WOS for developing mammary cancer. Previous studies have identified a looping interaction between a genomic region and a known cancer gene, PAPPA. However, the global role of three-dimensional organization in WOS is not known. Therefore, we generated Hi-C and RNA-seq data in rat mammary epithelial cells within and outside WOS. We compared the temporal changes in chromosomal looping to those in expression and find that interactions that have significantly higher counts within WOS are significantly enriched for differentially expressed genes that are higher in WOS. To systematically identify differential domains of interactions, we leveraged symmetric non-negative matrix factorization. This revealed clusters of dynamic regions that change their chromosome conformation between the two time points. Our results suggest that WOS-specific changes in 3D genome organization are linked to transcriptional changes that may increase susceptibility to breast cancer at an early age.

Speakers:

Brittany Baur

University of Wisconsin-Madison

04:00 PM - 04:20 PM (EDT)

CompMS - Selection of features with consistent profiles improves relative protein quantification in mass spectrometry experiments

Constructing gene co-expression networks is a powerful approach for analyzing high-throughput gene expression data towards module finding, gene function prediction, and disease-gene prioritization. While optimal workflows for constructing co-expression networks – including good choices for data pre-processing, normalization, and network transformation – have been developed for microarray-based expression data, such well-tested choices do not exist for RNA-seq data. Almost all studies that compare data processing/normalization methods for RNA-seq focus on the end goal of determining differential gene expression. Here, we present a comprehensive benchmarking of 30 different workflows, each with a unique set of normalization and network transformation methods, for constructing co-expression networks from RNA-seq datasets. We test all these workflows on both large, homogenous data (Genotype-Tissue Expression project) and small, heterogeneous datasets from various labs (submitted to the Sequence Read Archive). Our results demonstrate that choosing the between-sample normalization method has the biggest impact, with trimmed mean of M-values or upper quartile normalization producing networks that most accurately recapitulates known tissue-naive and tissue-specific gene functional relationships. Furthermore, we provide insights as to when other methods should be used and which experimental factors, including sample size, noticeably affect network accuracy.

Speakers:

Kayla Johnson

Michigan State University

04:10 PM - 04:20 PM (EDT)

Vari - Calibrating variant-scoring methods for clinical decision making

Identifying pathogenic variants and annotating them is a major challenge in human genetics, especially for the non-coding ones. Several tools have been developed and used to predict the functional effect of genetic variants. However, the calibration assessment of the predictions has received little attention. Calibration refers to the idea that if a model predicts a group of variants to be pathogenic with a probability P, it is expected that the same fraction P of true positive is found in the observed set. This problem is relevant since poorly calibrated algorithms can be misleading and potentially harmful for clinical decision-making. We evaluated the calibration and the prediction performance of four predictors that directly furnish a probability as output (DANN, DeepSea, FATHMM-MKL and PhD-SNPg). Despite the fact that they show similar performances in terms of AUC, most of them are not well calibrated. We also tested two methods that provide only raw scores, applying calibration transformations to them (CADD and Eigen). We showed that CADD can be well-calibrated and can be used, after this transformation, as a probability score. Among the predictors tested, PhD-SNPg provides the best calibration without any transformation of the scores, while CADD is the best predictor after calibration.

Speakers:

Silvia Benevenuta

University of Torino

04:10 PM - 04:20 PM (EDT)

NetBio: Network Biology - Supervised prediction of aging-related genes from a dynamic context-specific protein interaction subnetwork

Human aging is linked to many diseases. Because the aging process is influenced by genetic factors, it is important to identify human aging-related genes. We focus on supervised prediction of such genes. Gene expression-based methods for this purpose study genes in isolation from each other. While protein-protein interaction (PPI) network-based methods account for interactions between genes' protein products, current PPI network data are context-unspecific, spanning different biological conditions. Instead, we analyze an aging-specific subnetwork of the entire context-unspecific PPI network, obtained by integrating aging-specific gene expression data and PPI network data. We are the first to propose a supervised learning method for predicting aging-related genes from an aging-specific PPI subnetwork. In a comprehensive evaluation, we find that: (i) using an aging-specific subnetwork yields more accurate aging-related gene predictions than using the entire context-unspecific network, (ii) using a dynamic aging-specific subentwork is superior to using all static aging-specific subnetworks, and (iii) predictive methods that we propose outperform existing methods for the same purpose. Our best method achieves impressively high accuracy of 90%-95% (depending on the measure), compared to 72%-80% by the next best method. Our method could guide with high confidence the discovery of novel aging-related genes for wet lab validation.

Speakers:

Qi Li

University of Notre Dame

04:10 PM - 04:20 PM (EDT)

MICROBIOME - Assembly results for second round of CAMI challenges

Speakers:

Alex Sczyrba

Bielefeld University

04:20 PM - 04:40 PM (EDT)

HiTSeq: High Throughput Sequencing - Terminus enables the discovery of data-driven, robust transcript groups from RNA-seq data

Motivation: Advances in sequencing technology, inference algorithms and differential testing methodology have enabled transcript-level analysis of RNA-seq data. Yet, the inherent inferential uncertainty in transcript-level abundance estimation, even among the most accurate approaches, means that robust transcript level analysis often remains a challenge. Conversely, gene-level analysis remains a common and robust approach for understanding RNA-seq data, but it coarsens the resulting analysis to the level of genes, even if the data strongly support specific transcript-level effects. Results: We introduce a new data-driven approach for grouping together transcripts in an experiment based on their inferential uncertainty. Transcripts that share large numbers of ambiguously-mapping fragments with other transcripts, in complex patterns, often cannot have their abundances confidently estimated. Yet, the total transcriptional output of that group of transcripts will have greatly-reduced inferential uncertainty, thus allowing more robust and confident downstream analysis. Our approach,

28th Conference on

Intelligent Systems for Molecular Biology

ISMB 2020

Welcome to the ISMB 2020 Virtual Conference Platform!

Welcome to ISMB 2020, the flagship meeting of ISCB!

Schedule

Elaine Mardis

Laurence Calzone

Eric Merzetti

Diana Széliová

Michael Reich

Cathy Wu

Ana Conesa

Richard Roberts

Oliver Stegle

Rita Colwell

Sanjeev Dahal

Suhaib Mohammed

Madelaine Gogol

Simone Marini

Jane Skok

Multiple Authors

Fairlie Reese

Pablo Moreno

Shunfu Mao

Jean-Christophe Lachance

Andrey D. Prjibelski

Multiple Authors

Ronghui You

Xi Wang

Rabie Saidi

Yang Zhang

Dimitri Perrin

Paul Francois

Bali Pulendran

Baraa Orabi

Xiangying Jiang

Shamsuddin Bhuiyan

Nicola Bordin

Robert Leaman

Multiple Authors

Aayush Grover

Karen Oconnor

Yoseph Barash

Diane Kovats

Thomas Lengauer

Jérôme Waldispühl

Étienne Fafard-Couture

Alex Vesztrocy

Dorothy Ellis

Shuzhao Li

Douglas Lauffenburger

Russ Altman

Anshul Kundaje

Michelle Brazas

Michelle Brazas

Tugce Bilgin

Seth Cooper

Matthew Hahn

Dimitar Vassilev

Athanasios Zovoilis

Javier De Las Rivas

Bruno Gaeta

Ginger Tsueng

David Hawkins

Nicola Mulder

Yin Chew

Xuan Wang

Ziynet Nesibe Kesimoglu

Dina Svetlitsky

Samuel Anyaso-Samuel

Bruno Gaeta

Javier De Las Rivas

Nicola Mulder

Tugce Bilgin

Roman Sarrazin-Gendron

Cath Brooksbank

Sana Badri

Daniel Lobo

Patricia Palagi