[Ml-stat-talks] Wed 9/21: Barbara Englehardt on genome-wide associations

David Mimno mimno at CS.Princeton.EDU
Sun Sep 18 20:55:59 EDT 2011


For our first ML talk of the year, we have Barbara Englehardt from Duke.
The talk will be this Wednesday (9/21) at 12:30 in CS 402.

Title: Genome-wide associations studies with complex phenotypes: How
statistics can help

Abstract: Genome-wide association studies (GWAS), or studies to
identify genetic variants that are associated with a particular
phenotype or disease, can be performed trivially using available
software given sufficient numbers of individuals and simple
quantitative or case-control phenotypes. However, when the phenotype
of interest is complex (e.g., multivariate, noisy, difficult to
measure directly), and these simple approaches are insufficient,
statistical models can help identify potential associations. I present
two GWAS on different complex phenotypes. The phenotype in the first
study was an acute, subjective response to amphetamine, which is
difficult to measure in humans; we used sparse factor analysis to
project a large and noisy set of questionnaire responses down to a
low-dimensional space of interpretable phenotypes. The goal of the
second study was to identify a genetic locus that regulates gene
expression differently in the presence of statins versus in the
presence of a control buffer. Here, we developed multiple models of
differential regulation and computed Bayes factors to identify such
genetic loci. In both studies, we identified loci with potential
associations that also have interesting implications in terms of the
mechanism of both phenotypes. This was joint work with Matthew
Stephens.

Barbara graduated from Stanford University and received her PhD
from the University of California, Berkeley, advised by Professor
Michael Jordan. She did postdoctoral research at University of Chicago
working with Professor Matthew Stephens. Interspersed among her
academic experiences, she spent two years working at Jet Propulsion
Laboratory, a summer at Google Research, and a year at 23andMe. She
received an NSF Graduate Research Fellowship, the Google Anita Borg
Memorial Scholarship, and the Walter M. Fitch Prize from the Society
for Molecular Biology and Evolution. Her lab studies the underlying
biological mechanisms of complex phenotypes and human diseases through
statistical modeling.


More information about the Ml-stat-talks mailing list