Analysis of Microarray Gene Expression Data

University Of Colorado At Denver-Downtown Campus, Denver CO

Investigators

Abstract

The PI will spend one academic year in residence at the Department of Molecular, Cellular, and Developmental Biology (MCDB) at the University of Colorado at Boulder to study molecular biology, and to perform research in microarray data analysis (MDA). The core expertise of the PI is in development of numerical methods for eigenvalue problems applicable, e.g., to the Principal Component Analysis (PCA) of large data sets, and in investigation and computation of principal angles between subspaces that are closely connected to the Canonical Correlation Analysis (CCA). The PI's team is developing massively parallel software for PCA and CCA that can be used for data clustering. Modern microarray data provide vast amounts of useful biological information, but their analysis is computationally challenging. Molecular biologists need fast, reliable, and advanced MDA tools to locate clusters of genes responsible for specific biological processes. The purpose of this proposal is to allow the PI to gain the needed background and hands-on experience and training in molecular biology to pursue research in MDA in interaction and collaboration with experts in the area. The PI will be hosted in Min Han's laboratory in the MCDB. The Min Han laboratory is involved in cutting-edge research on finding functional gene clusters responsive to the novel regulation of growth and development of an organism through the fatty acid (FA) signaling using Affimetrix microarray chip technology. The data will be used for case studies. The goal of the proposed research is to apply the PI's expertise in PCA and CCA for developing novel mathematical algorithms specifically tailored for microarray datasets, including the recently developed GeneChip tiling arrays. The PI will investigate new classes of spectral clustering and bi-clustering techniques. The PI will apply the acquired knowledge in molecular biology to experiment with the novel PCA and CCA methods on case studies using real microarray experiments data. The new knowledge gained by the PI will make it possible to integrate the students' thesis research with applications in computational molecular biology, enable the PI to supervise students' research more efficiently, and be used for new course development. If successful, the new PCA and CCA methods developed and implemented by the PI for MDA will contribute to microarray use for scientific discovery.