Exploring Massive Gene Expression Data With A Novel Statistical Notion-Liquid Association

University Of California-Los Angeles, Los Angeles CA

Investigators

Abstract

Underlying most methods for analyzing gene expression is the key notion of profile similarity. Genes with similar profiles are likely to form structural complex or to participate in common cellular processes. Yet whether two genes are co-expressed or not is often affected by activities of other genes and sometimes a positive correlation can be turned into a negative one. In this research, Liquid association (LA) is introduced as a method to quantify changes in correlation. A preliminary study showed that LA can help identify functionally related genes that have no profile similarity. This development is harbored on three fronts: biological motivation, large scale computation, and statistical theory. The specific aims of the LA system include (1) to identify a small list of master expression genes- genes that have more expressional influence than others; (2) to help understand how genes regulate each other through important metabolic intermediates; (3) to analyze the co-expression pattern for genes associated with a known pathway; (4) to help the functional prediction of unknown genes; and (5) to provide an expression network through master genes. Large scale gene expression profiling has become increasingly more popular in tumor research and many other biological areas. Microarrays enable the full genome mRNA measurement. They have been applied to monitor gene activities under various physiological or environmental conditions. Investigations on differential expression between normal and disease tissues or cell-lines have also led to the identification of genes with potential diagnostic and clinic values. The LA-system will help unearth the biological information hidden beneath many expression databases that are publicly accessible. The results will be available for free use by other academic researchers. This grant is made under the Joint DMS/NIGMS Initiative to Support Research Grants in the Area of Mathematical Biology. This is a joint competition sponsored by the Division of Mathematical Sciences (DMS) at the National Science Foundation and the National Institute of General Medical Sciences (NIGMS) at the National Institutes of Health.