The Computational Life Sciences Seminar Series (CLSSS) was launched in October 2017. This seminar series highlights research in the fields of bioinformatics and computational life sciences, fosters collaborations, and provides a broad audience for feedback on current research. The format is flexible (e.g., standard seminars, lightning talks, etc.) and open to faculty, students, and researchers throughout the University of Chicago community.
We welcome your thoughts and feedback as we develop and grow this series.
May 30, 2019
Personalized Medicine to the Bedside: Using Bioinformatics in Pharmacogenomics Translational Research
Presented by Keith Danahey, MS
Keith Danahey of the Center for Research Informatics and the Center for Personalized Therapeutics presents on the process of bringing translational research in genomics to patients at the University of Chicago, Northwestern University, and the University of Illinois at Chicago. Innovative measures include a custom laboratory information management system, Thermo Fisher TaqMan Open Array Assay validation, genomic translations of genotypes to star alleles to phenotypes, designing a physician-centered portal: The Genomic Prescribing System (GPS), delivery of clinical decision supports, Epic integration, multi-institution implementation, genomic data analysis tools, and data visualization techniques for publications. These methods have led to the advance of personalized medicine for patients throughout Chicago.
April 25, 2019
Deep learning for image reconstruction in medical imaging: challenges and opportunities
Presented by Greg Ongie, PhD
Deep learning has the promise to revolutionize the field of image reconstruction in medical imaging. For example, preliminary studies have shown that deep learning approaches could allow for a ten-fold increase in the speed of MRI acquisitions, or allow for x-ray CT imaging at half the conventional radiation dose without compromising image quality. However, the success of deep learning is predicated on access to a large, well-curated database of ground truth training images. In many MRI and CT imaging applications, ground truth training images are scarce or non-existent, which makes the extension of off-the-shelf deep learning solutions to these settings challenging. Furthermore, in clinical settings it is essential that learning-based reconstruction methods do not hallucinate or erase critical diagnostic features in the images (e.g., tumors). Finally, deep learning approaches must also be robust to confounding factors such as noise, poor calibration, or patient movement. This talk will discuss these and other limitations of current deep learning approaches for MRI and CT image reconstruction, and highlight some possible solutions.
March 28, 2019
Uncovering the Genetic Architecture of Complex Diseases
Presented by Matt Dapas, Northwestern University
Understanding the genetic architecture of complex diseases is a central challenge in human genetics. Disease risk loci identified in genome-wide association studies (GWAS) can account for only a small proportion of the genetic heritability for most complex diseases. By developing new analytical approaches that combined clinical, GWAS, and whole-genome sequencing data from women affected by a common endocrine disorder, polycystic ovary syndrome (PCOS), we were able to determine whether various genetic phenomena such as rare noncoding variation and genetic heterogeneity contribute to the disease. Our findings reveal novel insights into the genetic architecture of PCOS that should inform the way complex diseases are studied going forward.
November 29, 2018
What Is the Value of 3 Billion Genotype-Phenotype Associations in Half a Billion Individuals?
Presented by Hae Kyung Im, PhD, Assistant Professor, Department of Medicine
Dr. Im reviews what we have learned about the biology of complex traits after thousands of GWAS studies have been conducted and how we can translate this knowledge into actionable targets to improve the health of individuals.
September 27, 2018
The Landscape of Somatic Genome Rearrangements in Human Cancers
Presented by Lixing Yang, PhD, Assistant Professor, Ben May Department for Cancer Research
High-throughput sequencing technologies make identification of somatic genome rearrangements in cancer patients possible although many challenges remain. We developed a novel algorithm that can not only detect these events in high confidence, but also infer their underlying mechanisms. By analyzing over 2,000 whole-genome sequenced cancer patients across 40 tumor types, we found tremendous diversity in terms of frequencies and types of somatic rearrangements among patients and across tumor types. The somatic events depend more on non-homologous based mechanisms compared to germline rearrangements. We identified many novel cancer-driving gene fusions and validated several by in vitro and in vivo assays. Furthermore, we found that ~30% of the patients have massively rearranged chromosomes, many of which are associated with up-regulation of oncogenes.
May 24, 2018
Analysis and Visualization of Life Sciences Data
Presented by Teodora Szasz, Image Analysis and Data Visualization Specialist, Research Computing Center
Teodora Szasz presents a discussion of tools for visualizing, manipulating, and understanding life science research data from many image modalities, including CT, MRI, 3D Microscopy, and other techniques. Examples of analysis and visualization are presented, including segmentation, registration, and 3D reconstruction. Also discussed are use cases of deep learning for medical image analysis and different deep learning toolkits that can be used in the life sciences.
View the presentation:
Analysis and Visualization of Life Science Data
April 26, 2018
Large Scale Mental Health Measurement
Presented by Robert Gibbons, PhD, Blum-Reise Professor of Bioistatistics
Dr. Robert Gibbons discusses how innovations in adaptive testing have helped to optimize the measurement of mental health. These innovations provide an alternative to full scale survey administration that has historically been based primarily on subjective judgment and classical test theory. Computerized adaptive tests can be developed using multidimensional extensions of item response theory, estimates of items (e.g., difficulty, discrimination), and individuals (e.g., severity of depression). These tests efficiently identify suitable item subsets for each individual that allows different individuals to receive different symptom items that are targeted to their specific impairment level. The result is reduction in patient burden, elimination of clinician burden, and maximization of precision of measurement.
March 1, 2018
Development and Implementation of Anchored Multiplexed PCR NGS Assays (DNA and RNA) for Studying Somatic and Germline Variants
Presented by Erin Crowgey, PhD, Associate Director of Bioinformatics, Nemours Alfred I. DuPont Hospital for Children
Dr. Erin Crowgey presents on the diagnostic and prognostic value of studying pediatric malignancies. Chromosomal rearrangements leading to the generation of gene fusions are more common in pediatric malignancies compared to adults and possess diagnostic and prognostic value. Identification of novel gene fusions provides a means for patient stratification and the foundation for the development of targeted therapeutics. The innovations discussed include: the development of custom Next Generation Sequencing (NGS) gene panels for acute myeloid leukemia (AML), and a newborn screening assay that detects germline events in high-risk populations.
February 22, 2018
Gaining Functional Immune System Insights Through the Application of Machine Learning on Immunofluorescent Microscopy Images
Presented by Vladimir Liarski, MD, Assistant Professor of Medicine, Section of Rheumatology
A discussion with Dr. Vladimir Liarski about his work on classifying the anatomic structures of human kidney using digital images and machine learning. This research introduces a method to objectively segment and classify the these images using Python to detect glomerular and non-glomerular areas. The methodology used to train the program to classify tissue and the benefits for diagnosis and treatment are discussed.
January 25, 2018
Genomics and Microbiome, Interconnections to Health and Obesity Studies
Presented by Dr. Beatriz Penalver Bernabe, Postdoctoral Researcher, Surgery Department and Cesar Cardona, PhD Candidate, Biophysical Sciences Program
A presentation of research currently underway in Jack Gilbert’s laboratory looking at how genomics of the microbiome might be connected to health and obesity outcomes. Topics include an introductory overview of bioinformatics and DNA sequencing, as well as amplicon and metagenomics sequencing for microbial communities.
November 15, 2017
A Data Management System for X-ray Reconstruction of Moving Morphology (XROMM) Datasets: Applications for Biomechanics, Neuromechanics and Orthopedics Research
Presented by Richard Williams IV, Ph.D., Application Developer/Computational Scientist, Research Computing Center (RCC)
Explore the archiving, analysis, and publication of X-ray Reconstruction of Moving Morphology (XROMM) datasets with the XROMM Data Management (XDM) system developed by the Research Computing Center. The XDM system packages study/trial data and metadata collected at the XROMM facility, transfers the data to the Midway high performance computing cluster for analyses and long-term storage, and allows the study/trial data to be accessible via a searchable web interface, the XMA Portal. The power and promise of XROMM is demonstrated by biological research on a range of projects, including jaw kinematics during feeding in fish, pigs, and ducks, rib kinematics of breathing in lizards, limb kinematics of locomotion in humans, pigs, dogs, birds and bats, and muscle architecture dynamics in rats, turkeys, fish and birds. Join us and learn about opportunities for using the University of Chicago XROMM core facility.
October 18, 2017
Environmental and Clinical Data Commons Initiatives at UChicago: Overviews, Demos, and Opportunities
Presented by Michael Fitzsimons, PhD, Director of User Services, Center for Data Intensive Science (CDIS) and Zac Flamig, PhD, Postdoctoral Scholar, CDIS
Currently, there is a large amount of data available to researchers across a spectrum of disciplines. However, accessing and standardizing these data are a challenge and serve as an impediment to research. The University of Chicago is involved in several initiatives to build harmonized data commons that greatly improve the ability of the research community to utilize “Big Data”. The speakers will showcase some of these initiatives involving both clinical/genomics data (e.g., Genomic Data Commons) and environmental data (e.g., OCC Environmental Data Commons). Come learn about opportunities to utilize these data in your own research and/or develop software that interacts with the commons framework.
View the presentations:
CDIS Biomedical Data Commons
Open Commons Consortium Environmental Data Commons