Skip to main content

Precise Functional Genomics of Digital Organisms via Metabolic Analysis

Dr. Daniel Weise
University of Washington

Date:  Friday, October 14, 2005
Time: 11:00am-12:00pm
Place: 3105 Engineering

Abstract: Functional genomics ascribes functions to genes, usually via genomic mutation studies to determine how the presence or absence of a gene affects phenotype. The realm of evolved digital organisms, with its wealth of data and history, allows us to compute functional genomics in the other direction, from phenotype to genotype. We do this by employing metabolic analysis that assigns the phenotypic effects of each gene in a digital organism. Metabolic analysis is not only many times faster than mutation studies, but is also immune to most of the effects of epistatic interactions that plague mutation studies. In addition, metabolic analysis determines whether a gene affects a given phenotype through metabolic or through regulatory pathways. We show the utility, accuracy, and importance of metabolic analysis by directly comparing its output to prior published results that use mutational studies. We discovered two unexpected benefits of metabolic analysis: first, because it is much more detailed than phenotypic information, it provides deep insights into evolutionary trajectories; second, that it shows that far more complexity has been evolving than has been assumed.

Biography: Daniel Weise is an affiliate faculty member of the Computer Science and Engineering Department of the University of Washington , where he has been for the last year. During the preceding 12 years, he was at Microsoft Research leading groups on program analysis and automatic bug detection, along with a stint in the Office product group getting his ideas used in practice instead of just having them sit around as ink on paper or as magnetic domains on rotating rust. Microsoft Research stole him away from Stanford University were he was a member of the EE faculty for many years. His PhD and MS are from MIT where he spent 6 years in the Artificial Intelligence Lab. This is probably the most boring bio he has ever written, but having changed research focus from programming languages and compilers to digital organisms, none of the old jokes work anymore. (Actually, there is a subtle joke here, but you have to attend the talk to get it.)