Invited Speakers
 |
Geoffrey Charles Fox
Distinguished Scientist and Director Website: http://www.dsc.soic.indiana.edu |
Deep Learning for Biomedical and Science Time Series[video] |
|
Abstract: |
|
| Deep Learning has been applied to many time series and sequence to sequence mappings but in many areas, the best way forward is not clear. Probably Industry is in the leads with audio applications and ride-hailing. We discuss a few research examples including Covid daily data, solutions of ordinary differential equations, hydrology and earthquakes. We consider both recurrent neural networks (LSTM) and Transformer architectures. We stress the importance of representing dependence on time and region (such as City where Covid data measured) and propose a framework for this. We show how working with the industry consortium MLPerf, we may be able to establish best practices and help the community discover and apply ideas to new fields. | |
Biography: |
|
| Dr. Geoffrey Charles Fox received a Ph.D. in Theoretical Physics from Cambridge University, where he was Senior Wrangler. He is now a distinguished professor of Engineering, Computing, and Physics at Indiana University, where he is the director of the Digital Science Center. He previously held positions at Caltech, Syracuse University, and Florida State University after being a postdoc at the Institute for Advanced Study at Princeton, Lawrence Berkeley Laboratory, and Peterhouse College Cambridge. He has supervised the Ph.D. of 73 students and published around 1500 papers (over 540 with at least ten citations) in physics and computing with a hindex of 82 and over 38500 citations. He received the High-Performance Parallel and Distributed Computing (HPDC) Achievement Award and the ACM - IEEE CS Ken Kennedy Award for Foundational contributions to parallel computing in 2019. He is a Fellow of APS (Physics) and ACM (Computing) and works on the interdisciplinary interface between computing and applications. He is involved in several projects to enhance the capabilities of Minority Serving Institutions. He has experience in online education and its use in MOOCs for areas like Data and Computational Science. He is active in Industry consortium MLPerf. | |
 |
|
 |
Li Li
Vice President of Clinical Informatics at Sema4 |
Analysis of the Mount Sinai Health System EMR reveals racial disparities and identified prognostic factors in the COVID-19 hospitalized patient population[video] |
|
Abstract: |
|
| During the COVID-19 pandemic in the greater New York City, we have utilized comprehensive EMR data on five member hospitals from the Mount Sinai Health System and analyzed clinical characteristics for COVID-19 patients. We have demonstrated higher age-adjusted rates of SARS-COV-2 infection in African Americans and Hispanics compared to their overall population in the greater New York City area, and identified that outcomes for hospitalized COVID-19 patients showed racial parity after adjusting for other clinical factors. We have discovered older age, lower oxygen saturation, elevated respiratory rate, and elevated lab parameters (WBC, creatinine and ALT) as clinically relevant prognostic indicators for increased risk of mortality in hospitalized COVID-19 patients, and developed a prognostic model to predict the outcomes for hospitalized COVID-19 patients based on baseline clinical characteristics at admission. These findings could help to identify which COVID-19 patients are at greatest risk of a severe infection response and predict survival. | |
Biography: |
|
| Li Li, M.D., MS., is Vice President of Clinical Informatics at Sema4, leading the development of a common data model for electronic medical records (EMR), driving the development and improvement of clinical applications for reproductive diseases, newborn health, and deep imaging learning and aiming to advance novel diagnostics, therapeutics, and insights into both diseases and wellness. Also, Dr. Li is an Assistant Professor at the Icahn School of Medicine at Mount Sinai in the Department of Genetics and Genomic Sciences. Dr. Li was trained as both a physician and bioinformatician, with an M.D. in Clinical Medicine from the Dalian Medical School in China and an M.S. in Bioinformatics from Boston University, and has over 17 years of experience in both industry (Quest Diagnostics) and academia (Stanford University, UCSD) with a focus on EMR and multi-omics data. At graduate school, she developed RDOCK, which has become one of the most widely-used protein-protein docking software programs. She successfully established the precision medicine groundwork including deep phenotyping of Lyme Disease, categorizing subtypes of T2D for treatment stratification rather than one size fits all model, developing deep machine learning algorithms to improve disease prognostics, identifying key clinical risk factors from EMR to improve quality of care for patients, and developing biomarker diagnostic assays for acute rejection and tolerance of kidney transplantation. She has received four Young Investigator Awards from the American Society of Transplantation and the Transplantation Society. Her three international patents led to the creation of the start-up company Organ-I Inc., acquired by IMMUCOR Inc., and one patent of identification of drug targets to treat kidney fibrosis and chronic graft injury. Dr. Li has published extensively in the fields of precision medicine, bioinformatics, and clinical bioinformatics, with more than 80 peer-reviewed papers in journals including Nature Biotechnology, Science Translational Medicine, and PNAS. | |
|
|
 |
Jianhua Xing
Associate Professor |
Reconstruct cell phenotype transition dynamics from single cell data[video], [slides] |
|
Abstract: |
|
| A cell is a dynamical system that receives and processes extracellular and intracellular information then respond accordingly through a complex regulatory network. A paramount task in systems biology is to learn the governing equations defined by the network structure. Decades of research on GRN has demonstrated that this is an unfathomable challenge. Recent exciting developments in the single cell genomics enable measurements of transcriptome (x) and estimation of RNA velocity (instant time derivatives of transcriptome, dx/dt) of hundred thousands of cells at unprecedented spatiotemporal resolution, and open the door to sufficiently powered inference. Building upon a comprehensive framework, dynamo, we fully takes advantage of the novel metabolic labeling based single-cell RNA-seq (scSLAM-seq) to improve the estimation of RNA velocity. With dynamo, we also show that genome-wide vector field function can be accurately and efficiently reconstructed with the improved velocity estimations. We also developed a live-cell imaging platform that aims to cross-examine trajectories inferred from scRNaseq data. Our method of single cell vector field reconstruction thus contributes as a significant step towards the holy grail of learning the governing equations of any cellular dynamic processes. | |
Biography: |
|
| Jianhua Xing is an Associate Professor in the Computational and Systems Biology Department, School of Medicine, and an affiliated faculty member of Department of Physics, University of Pittsburgh. He is also an affiliated member of University of Pittsburgh Hillman Cancer Center. Dr Xing’s research uses mathematical/computational modeling in combination with quantitative measurements to study the dynamics and mechanics of biological processes. Recently his lab focuses on reconstructing information of cell phenotypic transition dynamics from live cell time-lapse images and snapshot high-throughput single cell data. | |
|
|
 |
Reid Thompson
Assistant Professor Website: https://scholar.google.com/citations?user=tENxs6QAAAAJ&hl=en |
Human leukocyte antigen susceptibility map for SARS-CoV-2[video], [slides] |
|
Biography: |
|
| Reid Thompson is an Assistant Professor in Radiation Medicine, Biomedical Engineering, and Computational Biology, and an affiliated faculty member in Medical Informatics and Clinical Epidemiology at Oregon Health & Science University. He is also a Research Physician at the VA Portland Healthcare System. Dr. Thompson’s research focuses on precision oncology, and in particular on genomic predictors of immunotherapeutic outcomes. Recently, his lab has extended its work to the COVID-19 pandemic. | |
|
|
 |
Jake Y. Chen
Professor Website: https://bio.informatics.uab.edu |
PAGER-CoV: An Online Curated Gene Signature Database Resource for Coronavirus Disease Functional Genomic Studies[video] |
|
Abstract: |
|
| In this work, we report the development of a new web-based database called “Pathways, Annotated-gene-lists, Gene Signatures Electronic Repository for Coronavirus” (PAGER-CoV), located at http://discovery.informatics.uab.edu/PAGER-CoV. PAGER-CoV aims to help biomedical researchers characterize COVID-19 and other related disease molecular biology studies through curated gene set knowledge on Coronavirus-related infection, inflammation, organ damage, and repair. We will describe how the data were curated and basic/advanced features for the resource, using a few real-world queries and case studies. | |
Biography: |
|
| Dr. Jake Y. Chen is the Chief Bioinformatics Officer at UAB Informatics Institute and a Professor of Genetics, Computer Science, and Biomedical Engineering. Previously, he was the founding director of Indiana Center for Systems Biology and Personalized Medicine. He has over 25 years of R&D experience in biological data mining and systems biology, with over 170 peer-reviewed publications. He is currently President-elect of the Midsouth Computational Biology and Bioinformatics Society and an elected fellow of the American College of Medical Informatics. He also serves on the editorial boards of BMC Bioinformatics and Journal of American Medical Informatics Association. In 2019, he was recognized by Deep Knowledge Analytics as one of the "Top 100 AI Leaders in Drug Discovery and Healthcare". | |
|
|
 |
Jun Wan
Assistant Professor Website: https://wanbioinfo.github.io/Wan-bioinfo/index.html |
Genetic spectrum and distinct evolution patterns of SARS-CoV-2[video], [slides] |
|
Abstract: |
|
| COVID-19 outbreak, as a global emergency, has resulted in over twelve million confirmed cases including over a half million deaths as of July 10, 2020 all over the world. To uncover featured variations among SARS-CoV-2 genomes, we analyzed approximately thirty thousand high-coverage and high-quality sequencing data against Wuhan-Hu-1 strain and four bat coronavirus sequences including RaTG13. We found four main mutually exclusive clusters of mutations which covered over 96% of worldwide samples and predominated in different countries and areas. These signature mutations had distinguished temporal evolution patterns. Some cases were noticed to carry multiple different groups of SARS-CoV-2 mutations, suggesting potential superinfections which we must pay particular attention to. We also observed a significant overlap between sites of nucleotide substitutions among SARS-CoV-2 genomes and sites of RaTG13 coronavirus sequences different from Wuhan-Hu-1. Our findings may bring more clinical knowledge and help better treatments on individual patients. These unique discoveries may also provide deep insights into viral prevention and vaccine designs in near future. | |
Biography: |
|
| Jun Wan is an assistant professor in bioinformatics at Department of Medical and Molecular Genetics, Indiana University School of Medicine. He also directs the Collaborative Core for Cancer Bioinformatics (C3B) shared by NCI-comprehensive cancer center, Indiana University Simon Comprehensive Cancer Center, and NIC-designated Purdue University Center for Cancer Research. His research interests cover broad topics of bioinformatics and computational systems biology, particularly focused on gene regulatory network and epigenetic regulations. | |
|
|
 |
Jörg Menche
Principal Investigator Website: https://menchelab.com/ |
VRNetzer: A Virtual Reality Network Analysis Platform[video] |
|
Abstract: |
|
| An important chapter in the success story of network science is their capability of representing complex systems in a simple and intuitive fashion. Network visualizations play a key role in this context and can help generate new insights and hypotheses. However, the size and complexity of many real-world networks render common, two-dimensional visualizations impractical given the fundamental limitations in the amount of information that can be displayed on a relatively small screen or piece of paper. To overcome these limitations and unlock the full potential of networks, we have developed an immersive Virtual Reality (VR) platform that allows us to visualize and explore large networks. In my talk I will give a live demo of our VRNetzer and present our vision of an entirely new kind of data exploration platform in which human intuition can work seamlessly together with state-of-the-art analysis methods for large and diverse data. As an example, I will show an interactive exploration of genome-scale molecular networks for identifying genetic aberrations that cause rare diseases. The molecular network serves as an interface to commonly used public databases, allowing the user to retrieve detailed information on individual proteins or protein clusters, for example concerning their biological function or known disease associations. These databases can be used to annotate, filter and highlight specific portions of the network that are most relevant to a specific disease of interest or an individual patient. This enables researchers to quickly filter genomic variants from thousands to just a handful of promising candidates and at the same time helps developing hypotheses on their respective pathobiological mechanism that can be experimentally tested. Our platform is not limited to the biomedical test case shown in this presentation, but represents a general-purpose platform for arbitrary networks, annotations and computational methods applied on them. | |
Biography: |
|
| Jörg Menche studied physics in Leipzig, Recife and Berlin. During his PhD at the Max Planck Institute of Colloids and Interfaces in Potsdam (DE) he specialized in network theory. He then moved to Boston (US) to work as a postdoctoral at Northeastern University and at the Dana Farber Cancer Institute. In 2015, he started his own research group at the CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences in Vienna (AT). He was appointed full professor at the University of Vienna in 2020. His interdisciplinary team combines backgrounds ranging from biology and bioinformatics to medicine, physics, mathematics & arts. The broad ambition of his group is use tools and concepts from network theory to elucidate the complex machinery of interacting molecules that constitutes the basis of (patho-)physiological states. Major areas of interest are network-based approaches to rare diseases, understanding the basic principles of how perturbations of biological systems influence each other and developing novel Virtual Reality (VR) based technologies for analyzing large genomic data. | |