SISTA Seminar - Orly Alter
Start date: 8/06/2012
Location: ESAT 00.57
USTAR Associate Professor of Bioengineering and Human Genetics
Scientific Computing and Imaging (SCI) Institute
University of Utah
In the Genomic Signal Processing Lab at the University of Utah, we develop generalizations of the matrix and tensor computations that underlie theoretical physics, and use them to create models that compare and integrate different types of large-scale molecular biological data. We use our models to computationally predict physical, cellular and evolutionary mechanisms that govern the activity of DNA and RNA. Previous experimental results verify our computational prediction, demonstrating that mathematical modeling of DNA microarray data can be used to correctly predict previously unknown biological modes of regulation. We believe that future discovery and control in biology and medicine will come from the mathematical modeling of large-scale molecular biological data, just as Kepler discovered the laws of planetary motion by using mathematics to describe trends in astronomical data.
Our recent generalized SVD (GSVD) modeling of patient-matched data from The Cancer Genome Atlas (TCGA) draws a mathematical analogy between the prediction of cellular modes of regulation and the prognosis of cancers, and suggests that mathematical models created from cancer genomic data can be used to correctly predict the outcome of cancers . The GSVD and our recent higher-order GSVD (HO GSVD)  are the only algorithms to date that enable comparison of multiple patient-matched but probe-independent data. Ultimately we hope to bring physicians a step closer to one day being able to predict and control the progression of cancers as readily as NASA engineers plot the trajectories of spacecraft today.
1. Lee,* Alpert,* Sankaranarayanan and Alter, PLoS One 7, article e30098 (2012);http://dx.doi.org/10.1371/journal.pone.0030098
2. Ponnapalli, Saunders, Van Loan and Alter, PLoS One 6, article e28072 (2011);http://dx.doi.org/10.1371/journal.pone.0028072