Professor Joseph Sventek
Despite having sequenced the human genome over fifteen years ago, much is still unknown about how it functions. With the advent of high-throughput genomics technologies, it is now possible to measure properties of the genome across the entire genome in a single experiment, such as measuring where a given protein binds to the DNA or what genes are expressed. However, the complexity and massive scale (billions of base pairs with thousands of measurements each) of these data sets pose challenges to their analysis.
My research focuses on the development of new machine learning methods that address the challenges posed by genomics data sets. First, I will present a strategy for regularizing structured probabilistic models inspired by recent advances in graph-based semi-supervised learning. I will show how applying this strategy to genomics data sets in human cells allowed us to discover a new class of regulatory domain. Second, I will present a method for choosing which genomics experiments are most cost-effective to perform using on submodular optimization. Submodular optimization is a discrete analogue to convex optimization, which has revolutionized statistics and machine learning over the past few decades, yet submodular optimization is much less widely applied.
Maxwell Libbrecht is currently a postdoctoral fellow in Bill Noble's group at the University of Washington. He received his Ph.D. in 2016 from the Computer Science and Engineering department at University of Washington, advised by Bill Noble and Jeff Bilmes. He received his undergraduate degree in Computer Science from Stanford University, where he did research with Serafim Batzoglou. His research focuses on developing machine learning methods applied to genomics. He was the first author of a paper named one of ISCB's Top 10 Regulatory and Systems Genomics papers of 2015.