ISMB 2014 Special Sessions
Attention Conference Presenters - please review the Speaker Information Page available here.
ISMB 2014 features Special Session presentations throughout the conference July 13 - July 15. Special Sessions have the purpose of introducing the scientific community to relevant scientific issues and topics that are typically not within the focus of the conference. Preliminary program information on Special Sessions is noted below. (Schedules subject to change)
Room: Ballroom C
Organizer(s):
Dr. Niranjan Nagarajan is a Group Leader in the Genome Institute of Singapore and Adjunct Assistant Professor in the Department of Computer Science at the National University of Singapore. His research focuses on algorithmic and statistical issues in the study of infectious and genetic diseases. Dr. Nagarajan received a B.A. in Computer Science and Mathematics from Ohio Wesleyan University in 2000, a Masters in Computer Science from Cornell University in 2004 and a Ph.D. in Computer Science from Cornell University in 2006. He did his postdoctoral work in the Center for Bioinformatics and Computational Biology at the University of Maryland working on problems in genome assembly and metagenomics.
Mihai Pop is an Associate Professor in the Department of Computer Science and the Center for Bioinformatics and Computational Biology at the University of Maryland, College Park. He received his Ph.D. from Johns Hopkins University in 2000 and has been a Bioinformatics Scientist at The Institute for Genomic Research (TIGR) until 2005 when he joined the University of Maryland. Dr. Pop’s research focuses on computational analyses of genomic data (primarily sequencing and mapping data) with specific applications in sequence assembly, sequence alignment, and metagenomics.
Studies of the human microbiome have greatly expanded in size and scope in recent years. In particular, recent studies have highlighted the importance of gut microbiota in development (e.g. the development of the brain and the immune system) and disease (e.g. metabolic and colorectal diseases). Correspondingly, there has been an increased interest in establishing sophisticated computational techniques for studying microbial consortiums. In particular, analytical tools for modelling the interactions within and the dynamics of microbial communities are still in their infancy and are key to a systems-level understanding of the microbiome. The aim of this session is to foster greater interaction in this area between microbiologists and systems biologists and to seed scientific discussions through talks from 4 of the leading experts in this field.
Dr. Ehrlich is one of the leaders of the European MetaHit project – an effort to characterize the human gut microbiome and its role in health and disease. He will bring a broad outlook on the biomedical questions of current importance in the field.
Exploration of the human microbiome can now be conducted by massive DNA sequencing.An approach we name quantitative metagenomics provides a detailed view of the structure of microbial communities. One of its central hurdles is handling big data – typically some 50 million short DNA sequences are generated for each of the hundreds of samples handled in a routine study; they are used to monitor presence and abundance of 3.3 to 10 million microbial genes found in the gut communities (Qin et al. Nature, 464, 59-65, 2010; Li et al. Nature Biotechnology, in press). Gene profiles are thus generated for each sample. The profiles are correlated with the metadata related to human health status and/or life style by appropriate statistical approaches. The analytical space can be greatly reduced and the signal de-noised by clustering genes that are carried by the same genetic elements via abundance co-variance; about a half of the microbial genes can be assigned to some 8000 clusters representing bacterial species, phages, plasmids, CRISPR elements etc (Nielsen et al. Nature Biotechnology, in press). Using these approaches we have found that one of four people has low gut microbial richness and harbors a less healthy microbiome (Le Chatelier et al., Nature 500, 541-546, 2013). The loss of richness is correlated with a higher risk to develop metabolic syndrome associated pathologies such as type 2 diabetes, hepatic and cardio-vascular complications. The loss can be detected accurately and corrected, at least in part, by a nutritional intervention, in parallel with amelioration of metabolic parameters (Cotillard et al, Nature 500, 585-588, 2013). We thus seem to be in a position to detect the risk and act to alleviate it and possibly delay advent of common chronic diseases.
Dr. Turnbaugh is one of the world leaders in the study of human gut bacteria through computational and experimental models. He will provide a unique perspective on the experimental models used in this field and the associated computational needs.
Our gastrointestinal tracts harbor complex microbial communities (the gutmicrobiota/microbiome) that encode a vast array of enzymatic activities, contributing to the metabolism of our diet and the drugs we take. Yet the molecular mechanisms responsible often remain unknown, making it challenging to translate these findings to new therapies and diagnostics, or to appreciate the broader biological, ecological, and evolutionary implications. We are using a combination of computational and experimental approaches to elucidate these interactions between gut microbes and xenobiotics, including host-targeted drugs and diet-derived bioactive compounds. I will discuss two recent projects in the lab related to: (i) the impact of host-targeted drugs and antibiotics on the human gut microbiome, and (ii) the bacterial inactivation of the widely used cardiac drug digoxin. Ultimately, we aim to obtain a more comprehensive view of human metabolism, yielding fundamental insights into host-microbial interactions, and supporting translational efforts to predict and manipulate the metabolic activities of our resident gut microbes.
Dr. Bork is a prominent bioinformatician who has made significant contribution to the field of metagenomics, focusing on a broad range of questions including the persistence of microbial strains and the spread of antibiotic resistance in human microbiomes. He will provide a broad outlook on the
bioinformatics challenges faced by the field.
The human microbiome, that is all the microbes living in and around us, has recently becomeaccessible by environmental shotgun sequencing (metagenomics), whereby illumina technology allows large cohorts to be studied (Qin et al., Nature 2010). The most prominent human habitat of our invisible microbial companions is the gut, harboring in the order of 1000 species that are associated with important functions but also with more than 30 human diseases. I will illustrate the diagnostic potential of microbial markers using colon cancer as an example, but also illustrate the challenges that not only include sample heterogeneity and robustness as well as data handling, but also a variety of confounders. While metagenomics starts to deliver on differences between microbial abundances in several diseased compared to healthy individuals, the variation within the human population is still poorly understood. We recently identified three microbial community types at the genus level across several western countries from three continents, which we dubbed enterotypes (Arumugam et al, Nature, 2011). I will use this stratification of the human population to discuss methodological challenges in digesting complex data. We also analyzed (meta) genomic variation in gut microbial communities at the strain level and found that these variation patterns could serve as a fingerprint of an individual (Schloissnig et al., Nature, 2013). Despite the ongoing metagenomic data deluge, this level of resolution calls for further increased cohort sizes and improved metagenomic resources as currently only abundant strains can be sufficiently analysed.
Dr. Alm is a well recognized expert in the application of computational methods for studying the structure and evolution of microbial systems. He will provide an evolutionary biology perspective of microbial systems and the associated analytical challenges.
This part will delve deeper into computational approaches for studying the structure and
evolution of microbial system, particularly centered around the challenges related to the
analysis of quantitative microbial surveys.
Room: 312
Organizer(s):
Christine Orengo is a Professor in Bioinformatics at UCL, London. She is a Vice president of ISCB and Co-Chair of COSI task force. Christine is also Chair of the ISCB Conferences Committee.
Her research interests include protein structure classification and prediction. She is also interested in the evolution of protein functions and develops methods for predicting protein functions and protein networks. Her group collaborate with several experimental groups analysing funtional genomics data and metagenomics data. She is a member of EMBO and is involved in scientific advisory boards for SIB, TGAC and the BBSRC. See Website
Co-Chair of the COSI task force.
ISCB is keen to promote virtual networks and make the society a hub of electronic activity throughout the year, by setting up web communities that can share information, hold meetings and discuss ideas ‘long distance’ via the internet. These Communities of Special Interest (COSIs) will be built around major research themes within computational biology or important activities such as networks of training, mentoring or support. They initially involve groups of people who have already been organising themselves and holding Special Interest Group (SIG) meetings or workshops at the annual meeting of ISCB - ISMB.
Twelve COSIs are being launched this year and representatives from each will describe the activities of the community and highlights of their activities over the past and coming year.
ISCB plans to nurture these communities and has begun by hosting a web-portal that provides access to dedicated websites for each community. ISCB is supporting the computational infrastructure for the COSI websites using Wikimedia tools and there are future plans to video talks from COSI meetings/workshops to host on the COSI website.
We hope that you will support the COSIs and let us know of other communities who might want to become a COSI in the future.
Christine Orengo is Co-Chair of the COSI task force. She will give a brief overview of the history of the COSIs and outline plans for the future.
3D-SIG - focuses on structural bioinformatics and computational biophysics. http://cosi.iscb.org/wiki/3DSIG:Home.
CAMDA - Critical Assessment of Massive Data Analysis presents a crowd sourcing and open-ended data analysis challenge format which focuses on big heterogeneous data sets that are increasingly produced in several fields of the life sciences. http://cosi.iscb.org/wiki/CAMDA:Home.
AFP – Automated Function Prediction brings together computational biologists, experimental biologists and biocurators who are dealing with gene and gene product function prediction, to share ideas and create collaborations. http://cosi.iscb.org/wiki/AFP:Home.
VarI – Variant Interpretation promotes the formation of a collaborative network of scientists interested in the understanding of the meaning of genomic variation as applied to a range of questions, including population studies, functional and evolutionary impacts, and disease. http://cosi.iscb.org/wiki/VarI:Home.
HitSeQ – High Throughput Sequencing Algorithms and Applications, devoted to the latest advances in computational techniques for the analysis of high-throughput sequencing data including novel algorithms, analysis methods and applications in biology where high-throughput sequencing data has been transformative. http://cosi.iscb.org/wiki/HiTSeq:Home.
CompMS - Computational Mass Spectrometry - promotes the efficient, high quality analysis of mass spectrometry data through dissemination and training in existing approaches and coordination of new, innovative approaches. http://cosi.iscb.org/wiki/CompMS:Home.
RegSys – Regulatory and Systems Genomics focuses on computational methods that are important in the study of regulation of genes and systems. http://cosi.iscb.org/wiki/RegSIG:Home
NetBio – Network Biology serves to introduce novel methods and tools, identify best practices and highlight the latest research in the growing and interconnected field of network biology. http://cosi.iscb.org/wiki/NetBio:Home.
Bio-Ontologies - covers the latest and most innovative research in the application of ontologies and more generally the organisation, presentation and dissemination of knowledge in biomedicine and the life sciences. http://cosi.iscb.org/wiki/Bio-ont:Home
BOSC - The Open Bioinformatics Foundation (OBF) is a non-profit, volunteer-run group dedicated to promoting the practice and philosophy of Open Source software development and Open Science within the biological research community. http://cosi.iscb.org/wiki/OBF:Home.
Bioinfo-Core - is a worldwide body of people that manage or staff bioinformatics cores within organizations of all types including academia, academic medical centers, medical schools, biotechs and pharmas. http://cosi.iscb.org/wiki/Bioinfo-core:Home.
CoBe – Computational Biology Education focuses on bioinformatics and computational biology education and training across the life sciences. http://cosi.iscb.org/wiki/CoBE:Home.
JPI - Junior Principal Investigators aims to provide support during this process via a community of peers. http://cosi.iscb.org/wiki/JPI:Home.
Organizer(s):
Michael Waterman, University of Southern California, United States
Sorin Istrail, Brown University, United States
Room: Ballroom C
Organizer(s):
Steven E. Brenner is a Professor at the University of California, Berkeley, and also holds
appointments at Lawrence Berkeley National Laboratory and at the University of California, San
Francisco. As an undergraduate he studied in Walter Gilbert’s laboratory at Harvard College. He
received his M.Phil from the Department of Biochemistry at Cambridge University, and obtained
a Ph.D. from the MRC Laboratory of Molecular Biology and Cambridge where he studied with
Cyrus Chothia. After a brief visiting postdoctoral position at the National Institute of Bioscience
in Japan, he undertook postdoctoral studies under the mentorship of Michael Levitt. Brenner’s
research is primarily in the area of computational genomics, covering topics in protein structure,
RNA regulation, function prediction, metagenomics, and individual genome interpretation. He is
Founding Chair of the Computational Biology graduate program at Berkeley. He is currently a
director of the Human Genome Variation Society, and is a founding editor of PLoS Computational
Biology. He has served two terms as a director of the ISCB and was a founding director of the
Open Bioinformatics Foundation. His recognitions including being a Miller Professor, a Sloan
Research Fellow, a Searle Scholar, an AAAS Fellow, and named the recipient of ISCB’s Overton
Prize.
Mark Gerstein is the Albert L Williams professor of Biomedical Informatics at Yale University. He is co-director the Yale Computational Biology and Bioinformatics Program, and has appointments in the Department of Molecular Biophysics and Biochemistry and the Department of Computer Science. He received his AB in physics summa cum laude from Harvard College and his PhD in chemistry from Cambridge. He did post-doctoral work at Stanford and took up his post at Yale in early 1997. Since then he has published appreciably in scientific journals. He has >400 publications in total, with a number of them in prominent journals, such as Science, Nature, and Scientific American. (His current publication list is at http://papers.gersteinlab.org .) His research is focused on bioinformatics, and he is particularly interested in large-scale integrative surveys, biological database design, macromolecular geometry, molecular simulation, human genome annotation, gene expression analysis, and data mining.
Yu (Brandon) Xia is an Associate Professor of Bioengineering at McGill University. He graduated from Peking University with B.S. in Chemistry (major) and Computer Science (minor). He received his Ph.D. in Chemistry from Stanford University specializing in computational structural biology, and carried out postdoctoral research in protein bioinformatics at Yale University. His current research in computational biology and bioinformatics aims to construct genome-scale computer models of biomolecular networks with high spatial and temporal resolutions, and to use these genome-scale models to probe physical and design principles of biological networks, and to study the systems biology of disease.
Andrea Scaiewicz is a chemist by training. She received her Master’s Degree in Synthetic Organic Chemistry from the Hebrew University of Jerusalem, Israel. Her PhD, also from the Hebrew University, was focused on Structure-based drug design by Iterative Stochastic Elimination. As a grad student, she developed computational programs for fragment-based drug design based on combinatorial library optimization. She has led a chemo-informatics effort to sort databases into drug-like and non-drug-like molecules. Since 2010, she is a postdoctoral researcher with Prof. Michael Levitt at Stanford University. Her current research is focused on the nature of the protein universe. She develops methods that improve the ability to recognize and model protein sequences.
This session contributes to a celebration of Michael Levitt and his award of the Nobel Prize in Chemistry for foundational work in Computational Biology. Michael Levitt he has continually made research impact with work done by his own hands throughout computational biology. He has also been a mentor cherished by numerous trainees, many of whom now have successful academic careers. Michael Levitt has been a particularly generous, in every regard, in nurturing discovery and careers.
A goal of this session is to appreciate Michael Levitt’s role as an advisor through presentations from several of those who benefited from his mentorship. Levitt encouraged members of his group to work with tremendous independence, and so the breadth and extent of contributions through guiding his students and postdocs is relatively unappreciated. This session will help to give attendees a greater understanding of the diversity of research that Michael Levitt helped promote.
Enoch S. Huang received an AB in Molecular Biology from Princeton University (undergraduate advisor: Prof. Michael Hecht) and a PhD in Structural Biology from Stanford University, where he was a National Science Foundation Pre-doctoral Fellow in the laboratory of Prof. Michael Levitt. He was appointed a Jane Coffin Childs Fellow at Washington University School of Medicine (St. Louis), where he developed methods for protein structure prediction with Prof. Jay Ponder. In 1999, Enoch joined Cereon Genomics as a Computational Biologist. The following year, he accepted a position at Pfizer R&D in Cambridge as a Senior Research Scientist. In 2001, he became department head of the newly formed Molecular Informatics group and joined the site management team. In 2007 he accepted a global role as Head of the Computational Sciences Center of Emphasis.
External to Pfizer, Enoch has been an Adjunct Assistant Professor of Bioinformatics at Boston University since 2001. He currently serves on the Editorial Advisory Board for Drug Discovery Today, the Bioinformatics Professional Advisory Committee at Brandeis University, and the Industry Advisory Board of the International Society for Computational Biology. He has also served on the external advisory board of the Bioinformatics Program at the Rochester Institute of Technology, the program committee of the Systems Biology discussion group at the New York Academy of Sciences, the Steering Group for the Life Sciences Informatics Committee of the Massachusetts Biotechnology Council, and on Special Emphasis Panels of study sections for the National Institutes of Health. He is the author of over 30 research articles, scientific reviews, and book chapters and released the Open Source software package PFAAT.
Marie Brut received her Ph.D. from the Department of Nanophysics at the University of Toulouse, France, in 2009, and has been an associate professor of that department since 2011. Meanwhile, she was a postdoctoral researcher with Prof. Michael Levitt at Stanford University. During this period, she adapted her Ph.D. work, which aimed to develop computational programs for induced fit accommodation, to the prediction of large-scale motions in motor proteins. Her research is now focused on the development of new methodologies dedicated to the screening of mutations in oncoproteins. She mainly works on Ras oncoproteins in collaboration with the University Cancer Institute of Toulouse to propose new therapeutic strategies.
Jerry Tsai is an Associate Professor in the Chemistry Department at the University of the Pacific. He was first introduced to structural biology as an UCLA undergraduate working with Fred Eiserling on the genetic regulation of bacteriophage T4 tail length. Jerry earned his PhD under Michael Levitt at Stanford University in the Biophysics Program working on computational studies involving data mining molecular dynamics simulations. He continued as a post-doctoral fellow at the University of Washington with David Baker developing protein structure prediction methods. His current work focuses on developing an amino acid code for protein structure based on the simple 4 residue knob-socket motif.
Organizer(s):
Rachel Karchin, Ph.D. is an Associate Professor at Johns Hopkins University in Baltimore, MD. She is in the Department of Biomedical Engineering and Institute for Computational Medicine, with adjunct appointments in Computer Science and Human Genetics. She has developed bioinformatics- and machine-learning approaches to predict the impact of rare, inherited alleles in breast cancer and somatic mutations in tumor exomes, and is currently working on methods to identify prognostic biomarkers in pancreatic and liver cancers, and biomarkers of patient sensitivity to targeted cancer drugs
Melissa Cline, Ph.D. is a Project Scientist at UC Santa Cruz. She is affiliated with the cancer genomics group, where she analyzes functional cancer genomics data for TCGA and other projects. She is also affiliated with the Cancer Genomics Hub (CGHub), currently the world’s largest repository of cancer genomics data, where she manages sequencing data for cancer projects including TCGA, TARGET, CGCI and CCLE.
We are now in an era of unprecedented accumulation of data about the molecular basis of cancer. DNA sequencing of large cohorts of cancer patients has great potential to contribute to medical diagnostics, prognostics, and selection of individualized therapeutic regimens. We believe that bioinformatics analysis is essential for interpretation of the very large and high-dimensional data sets that are now emerging from the Cancer Genome Atlas, International Cancer Genome Consortium, TARGET, PCGP and others. The computational biology and genomics community has an opportunity to make a real difference in outcomes for cancer patients.
The purpose of this session is to help bioinformatics scientists understand better the current state of genomics and clinical decision making in cancer. We will have talks from physician scientists who are using genomics in their own decision making or who have taken a leadership role in making genomics-based decision making available to oncologists and their patients. We will discuss the needs of those in the clinical world, provide suggestions as to how computational scientists can get involved, and address some challenges involved when clinicians and computationalists work together. More information is available here.
Associate Professor, Department of Oncology and Otolaryngology-Head and Neck Surgery, Johns Hopkins University School of Medicine, Baltimore, MD, United States. Dr. Chung is a Director of Head and Neck Cancer Therapeutic Program leading the translational and clinical research efforts in head and neck cancer at JHU. She is also a co-Chair of Correlative Science and Translational Research Program at Eastern Cooperative Oncology Group (ECOG) and Radiation Therapy Oncology Group (RTOG) for the Head and Neck Cancer Committee. She has published 70+ publications in peer-reviewed journals including Cancer Cell, Journal of Clinical Oncology, New England Journal of Medicine and Journal of Clinical Investigation.
Head and neck squamous cell carcinoma (HNSCC) encompasses a diverse group of malignancies originating in the oral cavity, oropharynx, larynx and hypopharynx. Due to the heterogeneous nature of HNSCC, discovering a ‘‘silver bullet’’ or single critical biomarker to treat HNSCC, such as imatinib for EGFR-mutated chronic myeloid leukemia (CML), is unlikely. Many HNSCC patients now have the option for tumor profiling, using deep exome sequencing on panels of known driver genes. For patients with substantial financial resources, whole exome sequencing (WES) and whole genome sequencing (WGS) are possible. But, assuming that the patient lives long enough to receive such results, are the benefits to survival and quality of life worth the cost? This presentation will discuss application, potential benefits and current limitations of genomic technology in HNSCC care from perspectives of a practicing physician and patients.
Assistant Professor, Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX, United States. Dr. Parsons is the Director of the Pediatric Center for Personal Cancer Genomics and Therapeutics and Co-Director of the Cancer Genetics and Genomics Program and the Neuro-Oncology Program. He has published more than 25 scientific papers in peer-reviewed journals, including Science, Nature, and the New England Journal of Medicine. He has been awarded numerous honors for his research, including the Peter A. Steck Memorial Award for Brain Tumor Research (2009). Dr. Parsons is a Graham Cancer Research Scholar at Texas Children’s Cancer Center
Modern sequencing technologies can provide genome-scale data to oncologists and geneticists caring for cancer patients. But how well does sequencing technology aid patient care? The BASIC3 (Baylor Advancing Sequencing into Childhood Cancer Care) study is assessing the clinical impact of incorporating CLIA-certified exome sequencing into the care of children with newly diagnosed solid tumors. Tissue samples are submitted for clinical exome sequencing, with the results deposited into the patients’ medical records and disclosed to families by their oncologist and a genetic counselor. Oncologists are surveyed on treatment options in the event of recurrence before and after receiving sequencing results. Patients will be followed for two years. This presentation will discuss the clinical utility of exome sequencing, and the response of doctors and patients’ families to genetic testing.
Professor, Department of Biomolecular Engineering, University of California, Santa Cruz, United States. Dr. Stuart is the Associate Director of the Center for Biomolecular Science and Engineering for Cancer and Stem Cell Genomics, holds a Baskin Engineering Endowed Chair, and was an Alfred P. Sloan Fellow from 2005 to 2012. He is a co-leader of the TCGA Pan-Cancer network and a principal member of the Stand Up To Cancer Prostate Cancer Dream Team.
This talk will address some challenges bioinformatics scientists face in translating genomic data to a form useful for patient outcomes. First, cancer is often studied in silos defined by tissue of origin. However, recent pan-cancer analyses reveal connections across tissues, cell-of-origin signals, and contributions from the tumor microenvironment. Bioinformatics methods that deconvolute these signals may provide insights for treatment decisions. Next, a recent benchmark study revealed that current algorithms to identify DNA sequence changes in tumor genomes have limited concordance with each other. The DREAM somatic mutation calling challenge was recently launched to find the most accurate algorithms to read tumor genomes with high fidelity. Finally, algorithms are sorely needed to interpret the impact of these findings on the genetic pathways of cells. Approaches that can find explanatory models of tumor wiring and re-wiring may provide clues about what points in the networks can be targeted to eliminate or keep in check rogue cancer cells.
Organizer(s):
Dr. Kellis is a Professor of Computer Science at MIT and the Broad Institute and directs the MIT Computational Biology Group. He pioneered methods for the discovery of regulatory elements and circuits using comparative genomics, epigenomics, and human genetics, and he helps direct the integrative analysis efforts of several large-scale genomics projects, including the NIH Roadmap Epigenomics project, the ENCODE project, the comparative analysis of 29 mammals, and the Genotype Tissue-Expression (GTEx) project. Dr. Kellis received the US Presidential Early Career Award in Science and Engineering (PECASE), the NSF CAREER award, the Alfred P. Sloan Fellowship
Dr. Hirschhorn is the Professor of Pediatrics and Genetics at Harvard Medical School and the Broad Institute. He uses human genetics and genomics to understand the genetic basis of common diseases and quantitative traits, including obesity and height. He has organized a large international consortium to identify genes and variants that influence anthropometric measures of human body size and shape. His also uses genetic data to gain insights into the underlying biology of human body size and of other polygenic diseases and traits, genetic architecture, and population genetics. Other areas of interest include diabetic kidney disease and asthma.
Dr. Alkes Price is an assistant professor of statistical genetics in the departments of Epidemiology and Biostatistics at the Harvard School of Public Health, and an associate member of the Broad Institute. His research focuses on the development of statistical methods for uncovering the genetic basis of human disease, and on the population genetics underlying these methods. Areas of interest include components of heritability and genetic architecture of complex traits, disease mapping in structured populations, and methods for analyzing rare variant data.
Shamil Sunyaev is a Professor of Medicine at Harvard Medical School and Brigham & Women’s Hospital and an associate member of the Broad Institute. He is interested in evolutionary forces shaping population genetic variation. He studies spontaneous mutagenesis, functional effects of mutations and allelic variants, and population genetics. He develops computational and statistical methods for the analysis of DNA sequencing data and methods for predicting the effect of mutations and SNPs. on function using comparative genomics, protein structure, and functional genomics data. He combines evolutionary models and statistical methods for the analysis of complex phenotypes, and applies the knowledge gained to medical genetics studies and genetic studies on model organisms.
Understanding the architecture of complex human disease poses a number of computational and statistical challenges of key relevance to the ISMB community. One of the greatest surprises of genome-wide association studies over the last 10 years is the realization that disease-associated loci are predominantly non-coding, and that their effects are dramatically weaker than previously suspected. This has important implications in diagnosis, prevention, and treatment, and calls for new methods that integrate genetic association maps and functional genomics annotations of non-coding DNA elements from diverse tissues and cell types. This is made possible with the convergence of studies of common human variation, genetic association maps for several thousand human traits, and biochemical activity maps of functional non-coding DNA elements across hundreds of tissues and cell types.
This session presents novel statistical models for genetic and regulatory genomics integration, including gene-expression integration, protein-protein interactions, heritability estimation and partitioning, comparison of common and rare variants, comparison of coding and non-coding variants, integration of epigenomic annotations, inference of regulatory networks, and their use in recognizing regulatory pathways harboring weakly-associated non-coding variants. We expect these methods to have a profound impact on our understanding of complex human traits and human disease.
Dr. Hirschhorn is the Professor of Pediatrics and Genetics at Harvard Medical School and the Broad Institute. He uses human genetics and genomics to understand the genetic basis of common diseases and quantitative traits, including obesity and height. He has organized a large international consortium to identify genes and variants that influence anthropometric measures of human body size and shape. His also uses genetic data to gain insights into the underlying biology of human body size and of other polygenic diseases and traits, genetic architecture, and population genetics. Other areas of interest include diabetic kidney disease and asthma.
A challenge of genome-wide association studies of polygenic traits and diseases is to be able to use these loci to prioritize biologically relevant genes and pathways. Genome wide association studies of anthropometric traits have identified hundreds of associated loci. We have used existing and new computational methods to highlight known and novel genes and pathways as relevant to regulation of human height and body mass index. Our work shows that increasing the number of associated loci by increasing sample size increases the amount of biological information, and also that integrating genetic and other sources of information (including expression data and protein-protein interaction data) also increases the biological information that can be extracted from genome-wide association studies. We have implemented an integrative approach in a new tool, called Data-driven Expression Prioritized Integration for Complex Traits (DEPICT).
Dr. Alkes Price is an assistant professor of statistical genetics in the departments of Epidemiology and Biostatistics at the Harvard School of Public Health, and an associate member of the Broad Institute. His research focuses on the development of statistical methods for uncovering the genetic basis of human disease, and on the population genetics underlying these methods. Areas of interest include components of heritability and genetic architecture of complex traits, disease mapping in structured populations, and methods for analyzing rare variant data.
Common variants implicated by genome-wide association studies (GWAS) of complex diseases are known to be enriched for coding and regulatory variants. We applied methods to partition the heritability explained by genotyped SNPs across functional categories (while accounting for shared variance due to linkage disequilibrium) to imputed genotype data for schizophrenia and 10 other common diseases. We determined that non-coding functional annotations spanning a small fraction of the genome explain most of the genomic heritability. Our results highlight the value of analyzing components of heritability to unravel the functional architecture of common disease.
Shamil Sunyaev is a Professor of Medicine at Harvard Medical School and Brigham & Women’s Hospital and an associate member of the Broad Institute. He is interested in evolutionary forces shaping population genetic variation. He studies spontaneous mutagenesis, functional effects of mutations and allelic variants, and population genetics. He develops computational and statistical methods for the analysis of DNA sequencing data and methods for predicting the effect of mutations and SNPs. on function using comparative genomics, protein structure, and functional genomics data. He combines evolutionary models and statistical methods for the analysis of complex phenotypes, and applies the knowledge gained to medical genetics studies and genetic studies on model organisms.
The availability of large-scale genotyping and sequencing data opens a perspective for testing models of evolution, maintenance and allelic architecture of complex traits. Analytical approaches suggest that allelic architecture of complex traits is unlikely to be dominated by rare variants of large effects. Comparison of computer simulations with results of association studies may further constraint allelic architecture models, even though a large number of models remain consistent with the empirical results. We developed new computational methods for estimating the fraction of heritability due to common and rare allelic variants and variants in various functional classes including coding and regulatory variants. These methods can be applied to incoming large-scale sequencing data for phenotyped populations.
Manolis Kellis is a Professor of Computer Science at MIT and the Broad Institute and directs the MIT Computational Biology Group. He pioneered methods for the discovery of regulatory elements and circuits using comparative genomics, epigenomics, and human genetics, and he helps direct the integrative analysis efforts of several large-scale genomics projects, including the NIH Roadmap Epigenomics project, the ENCODE project, the comparative analysis of 29 mammals, and the Genotype Tissue-Expression (GTEx) project. Dr. Kellis received the US Presidential Early Career Award in Science and Engineering (PECASE), the NSF CAREER award, the Alfred P. Sloan Fellowship.
While the methodological framework for relating genotype directly to disease has been well established through the development of genome-wide association studies (GWAS), the incorporation of functional annotations, epigenomic information, and intermediate molecular phenotypes is still a great challenge. In this talk, I’ll describe some of our work seeking to address these challenges. (1) In the context of ENCODE and the Roadmap Epigenomics program, I will describe methods for integration of functional genomics datasets into chromatin state annotations, linking of regulators to enhancer regions and their target genes, and prediction and validation of causal regulators using massively parallel reporter assays. (2) I will present new methods for detection of weak-effect variants concentrated in regulatory regions based on their enrichment in specific tissues and cell types (3) I will present a new Bayesian method for integration of disease-associated variants in regulatory networks to predict additional disease-associated genes and fine-map individual regions based on their regulatory interactions. Overall, our results suggest 1000s of independent loci are associated with complex disease, concentrated in regulatory regions, and densely linked in gene regulatory networks.