Health Informatics Invited Speaker Series

The Invited Speaker Series is an extension of our Health Informatics Grand Rounds. In this series, top informaticians from around the country will share their current research with the University of Minnesota informatics community. This series is open to the public.

The series will generally be on Thursdays from 3:30-4:30, but dates and times may vary.

May 18-Deriving Novel Biological Insights from Existing Big Data: Computational Analyses of Cancer Epigenome and Transcriptome

May 18-Deriving Novel Biological Insights from Existing Big Data: Computational Analyses of Cancer Epigenome and Transcriptome

Picture of Wei Li
 
Wei Li, PhD
Professor of Bioinformatics in the 
Dan L. Duncan Cancer Center at Baylor College of Medicine

 

Time: 1:30 p.m. - 2:30 p.m. 

Location: 2-520 Moos Tow

Alternative polyadenylation (APA) is emerging as a pervasive mechanism in the regulation of more than 70% of human genes. The role of APA in human cancer is largely overlooked due to the lack of genome-wide APA profiling on large clinical cohorts. To overcome this limitation, we recently developed DaPars algorithm (Nature Communications 2014) for the de novo identification of APA from standard RNA-seq. Through DaPars analyses of thousands of TCGA tumor RNA-seq data, we revealed ~1,300 highly recurrent 3’UTR shortening genes, nominated CFIm25 (Nature 2014) as a master regulator of 3`-UTR shortening that links APA to glioblastoma tumor suppression, and suggested a surprising enrichment of 3’UTR shortening among transcripts that are predicted to act as competing-endogenous RNAs (ceRNAs) for tumor suppressor genes (Nature Genetics 2018). Our model-based analysis of the trans effect of 3ʹ-UTR shortening (MAT3UTR) reveals a significant role in altering ceRNA expression. MAT3UTR predicts many trans-targets of 3ʹ-UTR shortening, including PTEN, a crucial tumor suppressor gene involved in ceRNA crosstalk with nine 3ʹ-UTR shortening genes, including EPS15 and NFIA. These results underscore the power of innovative bioinformatics analyses that can derive novel biological insights from existing big data in cancer genomics.

 

 

 

 
 

May 16, 2018 - Population Health Issues and Opportunities for Data Science

May 16, 2018 - Population Health Issues and Opportunities for Data Science

Picture of Don Casey, MD, MPH, MBA

 

Don Casey, MD, MPH, MBA
Principal and Founder IPO4Health

Time: 1:30 p.m. - 2:30 p.m. 

Location: 2-690 Moos Tower 

Click for Recording

 Dr. Casey will present new perspectives on common clinical issues facing the healthcare delivery system and propose some thought provoking analytic approaches designed to solve these challenges.

 

April 11, 2018- Big Data and EHRs: Why is it so hard?

April 11, 2018- Big Data and EHRs: Why is it so hard?

Picture of Justin StarrenJustin Starren, MD, PhD, FACMI
Professor of Preventive Medicine and Medical Social Sciences at the Northewestern Feinberg School of Medicine

Time: 2:30 p.m. - 3:30 p.m. 

Location: 2-530 Moos Tower 

Click for Livestreaming 

 

 

 

 

 

 

Healthcare in this country has undergone a seismic transformation with the widespread adoption of Electronic Health Records (EHRs).  At the same time the growth of Genomics and Precision Medicine are reshaping fundamental assumptions about the nature of disease. Unfortunately, current generation EHRs is not designed for the big data challenges of Precision Medicine.  The Electronic Medical Records and Genomics (eMERGE) Project is a national consortium that has been a leader in mining EHR data to support research and also at developing methods for incorporating genomic data into care.  This talk will leverage the experiences of eMERGE and similar projects to illuminate the promise and challenges of bringing big data into healthcare.

March 21, 2018 - Structural Genomic Rationale for the Increased Incidence of Osteosarcoma in Canines

March 21, 2018 - Structural Genomic Rationale for the Increased Incidence of Osteosarcoma in Canines

Picture of Aaron Sarver

Aaron Sarver, PhD
Associate Professor, Institute for Health Informatics, Bioinformatics Research Scientist, Masonic Cancer Center, University of Minnesota

Time: 2:30 p.m. - 3:30 p.m. 

Location: 2-530 Moos Tower 

Click for Livestreaming 

 

 

 

 

 

 

Osteosarcoma (bone cancer) occurs at a very low rate in human children (~600 cases a year in the US) but occurs at much higher rates in adult large breed dogs (approaching 20% in certain breeds). To identify the molecular rationale for this observation, we used a comparative genomic approach to identify aberrations responsible for osteosarcoma etiology in a Sleeping Beauty transposon-accelerated mouse model, in human patients, and in naturally occurring canine tumors, using RNA-Seq and exome paired tumor normal analyses. We show that while a diverse assortment of conserved molecular mechanisms are present in tumors from all 3 species, the majority of dogs lose both copies of the PTEN tumor suppressor and provide evidence that this difference is due to the genomic location of PTEN within the canine genome, making them uniquely susceptible to PTEN loss. This result is fundamentally important, because it suggests that genetic engineering of the canine genome may reduce the increased risk of OS development. 

 

 

February 19, 2018 - The Inherited Genomics of Childhood ALL- from Biology to Clinical Implications

February 19, 2018 - The Inherited Genomics of Childhood ALL- from Biology to Clinical Implications

Picture of Jun Yang
Jun Yang, PhD
Associate Professor, Department of Pharmaceutical Sciences, Department of Oncology, and the Hematological Malignancies Program of St. Jude Comprehensive Cancer Center

Time: 3:30-4:30
Location: 2-690 Moos Tower
Recording

Abstract

Acute lymphoblastic leukemia (ALL) is the most common malignancy in children and also a prototype of cancer that can be cured by chemotherapy alone. Genomic profiling of ALL has revealed remarkable complexity of somatically acquired genetic lesions in this disease and novel therapeutic targets. However, the roles of inherited genetic variation in ALL pathogenesis remain poorly understood. Recent studies by us and others have comprehensively identified germline genetic variants associated with ALL risk, ranging from common polymorphisms related to modest ALL susceptibility to rare mutations driving familial predisposition. More importantly germline variants can also strongly influence host toxicity of ALL therapy, with direct implications in precision medicine in childhood cancers. In conclusion, inherited genetic variation is a critical part of ALL genomics, contributing significantly to inter-individual variability in ALL susceptibility, anti-leukemic drug response, and treatment toxicity.

January 25, 2018 - An Ontological Approach to EHR Data Quality

January 25, 2018 - An Ontological Approach to EHR Data Quality

Picture of Steven Johnson
Steven Johnson, PhD
Vice President of Technology and Clinical Informatics at Wolters Kluwer Health

Time: 3:30-4:30
Location: 2-690 Moos Tower
Recording

Abstract

Electronic Health Records (EHRs) are almost universally adopted in healthcare and now provide mountains of clinical data. There are high hopes that applying data science, machine learning and artificial intelligence (AI) to that data will help us achieve precision medicine, reduce the cost of healthcare and improve population health. But EHR data is often of poor quality and may not be sufficient to support these goals. 

In this talk, Dr. Johnson will discuss why we should care about data quality, how we can define it, measure it and ultimately, how we can improve it so that the data is fit for the purposes that we have for it.