The Department of Biochemistry & Molecular Genetics presents:
Gavin Ha, Ph.D.
Associate Professor
Herbold Computational Biology Program, Fred Hutch Cancer Center
Title: "Advanced machine learning methods to profile disease phenotypes from cell-free DNA"
Abstract:
Accurately diagnosing the disease phenotype is essential for clinical care. In cancer, changes in tumor histology and molecular subtype can emerge during treatment resistance. However, obtaining a biopsy to determine the tumor histology presents challenges for patients with advanced cancers.
Cell-free DNA (cell-free DNA) is a non-invasive 'liquid biopsy' solution that overcomes this limitation of tissue accessibility. Current cfDNA approaches primarily focus on detecting genomic alterations, but these alterations alone may not fully delineate disease subtypes or explain treatment resistance.
In this talk, I will present on methods that we have developed to infer transcriptional activity from the inference of nucleosome profiles in cfDNA. We leverage this approach to classify tumor subtypes in breast, prostate, and lung cancers with the goal of monitoring treatment-induced phenotype changes. I will also present on using this approach to predict the risk of developing preeclampsia in pregnant individuals. These studies showcase the potential to expand on the utility of cfDNA for precision medicine.
Host: Dr. Yaping Liu, PhD
Refreshments will be served.
Ashley Martin
Email