Monday, June 24, 2019
7:30 AM – 11:30 AM
For more information, please contact:
|7:30 AM||Registration and Breakfast|
Marylyn Ritchie, PhD
|8:10 AM -
Workshop: Genome Wide Association Study using Penn Medicine Biobank and the UK Biobank
|9:20 AM -
Workshop: Unlocking Study Variables from Clinical Notes using PennSeek and Linguamatics I2E
Nebo Mirkovic, PhD
|10:30 AM -
Workshop: Using PennChart Slicer/Dicer Tool for EHR Data Exploration
Genetic discovery experiments along with analysis of large datasets including genetic data and clinical data can be extremely useful. Penn Medicine BioBank is a terrific resource that includes over 50,000 consented participants from our health system. We have genome-wide genotype data and whole exome sequence on a large number of those participants. There is also a public resource (the UK Biobank), which includes over 500,000 individuals’ genetic and clinical data that can be used for research. In this presentation, we will describe these two resources including: how to gain access to these resources, what types of analyses can be done, and what tools we are building to facilitate easy access to the information in these EHR-linked biobanks.
For some research studies, a large proportion of study variables are “locked” in clinical notes. Natural language processing (NLP), a technology at the intersection of computer science, artificial intelligence, and computational linguistics, can be used to “unlock” study variables from clinical notes. We will explore two NLP tools – PennSeek and Linguamatics I2E – for understanding where study variables are documented across subspecialties and note types as well as how study variables can be extracted from clinical notes to support clinical and translational research.
One of the first steps to conducting clinical research is determining whether sufficient clinical data exists to study a patient population. Researchers may identify a patient population of interest based on their demographics, histories, diagnoses, and procedures among other characteristics. We will demonstrate how the PennChart Slicer/Dicer tool facilitates complex EHR data exploration and study cohort construction.