Transcriptional regulation is one of the crucial mechanisms used by living systems to regulate protein levels. Toxic effects of many chemicals are reflected in the dysregulation of gene expression, and gene expression changes are often reliable markers of a disease. Understanding the mechanisms of gene regulation is likely to improve our ability to effectively treat human disease and predict effects of environmental toxicants. Identifying groups of co-expressed genes by the cluster analysis of microarray data has been a commonly used approach for characterizing patterns of gene expression. Results of such analyses have often been used as a starting point for dissecting the regulatory mechanism driving the co-expression. Two examples of such approaches are identifying common putative regulatory motifs in cis-regulatory regions of such co-expressed genes and correlating patterns of co-expression with genomic events uncovered through the use of microarray based Comparative Genomic Hybridization. We propose to develop novel mathematical models and corresponding computational tools for efficient and reproducible extraction of relevant expression patterns, related regulatory motifs and genomic aberrations by jointly modeling genomic and functional genomic data. The proposed work will address the issue of developing a practical mathematical framework for an integrated analysis of different types of genomic and functional genomic data. Proposed computational procedures will be based on the context-specific Bayesian infinite mixture model. Joint modeling of genomic and functional genomic data will facilitate optimal information exchange between various data types. Proposed models will be validated by analyzing synthetic and real-world datasets. Corresponding computer programs will be freely distributed to the biomedical community. By using these programs, biomedical researchers will be able to make reliable and reproducible conclusions about gene expression patterns and associated regulatory mechanisms.

Agency
National Institute of Health (NIH)
Institute
National Library of Medicine (NLM)
Type
Small Research Grants (R03)
Project #
5R03LM008248-02
Application #
6887387
Study Section
Special Emphasis Panel (ZLM1-MMR-R (J2))
Program Officer
Sim, Hua-Chuan
Project Start
2004-04-19
Project End
2007-03-31
Budget Start
2005-04-01
Budget End
2007-03-31
Support Year
2
Fiscal Year
2005
Total Cost
$76,750
Indirect Cost
Name
University of Cincinnati
Department
Public Health & Prev Medicine
Type
Schools of Medicine
DUNS #
041064767
City
Cincinnati
State
OH
Country
United States
Zip Code
45221
Liu, Xiangdong; Jessen, Walter J; Sivaganesan, Siva et al. (2007) Bayesian hierarchical model for transcriptional module discovery by jointly modeling gene expression and ChIP-chip data. BMC Bioinformatics 8:283