|About GSBS | FAQ | Job Opportunities | Search UMDNJ|
Second Military Medical University
Thesis Advisor: Richard S. Nowakowski, Ph.D.
Graduate Program in Neuroscience
Wednesday, January 7, 2009
Gene co-expression is important in many physiological functions. However the mechanism of most gene co-expression events is not clearly understood. By employing microarray technology and including miRNA in addition to mRNA in gene expression analysis, we studied gene co-expression in mouse brain in an attempt to gain insight into the potential mechanisms. We found clusters of genes whose expression levels are inversely correlated with that of miRNA and clusters of genes whose expression levels are positively correlated with that of miRNA. In the inversely correlated clusters, by using transcription factor binding site and miRNA target prediction, we proposed that miRNA regulates expression of the co-expressed genes by down-regulating common transcription factors of these co-expressed genes. In those positively correlated clusters, transcription factor binding sites were found common to the miRNA and co-expressed mRNAs, therefore they are likely under the same same transcription factor control. There is also a possibility of a “double negative” mechanism in these clusters.
Next, we looked for genes that showed different expression patterns in C57BL/6J and DBA/2J. Gene ontology analysis showed these genes are enriched in functional groups such as those involved in in transcription regulation, RNA and protein modification, nervous system development and signal transduction, including neurotransmission. These genes could potentially explain some of the phenotypic differences between the two strains.
We applied the weighted gene co-expression network analysis to DBA/2J Affymetrix microarray data. Network modules that correspond to genes mainly showing higher (or lower) expression levels for specific brain regions were identified. By combining gene significance and intramodular connectivity, “hub genes” were also identified. These “hub genes” have both high gene significance and intramodular connectivity. The high gene significance explains their relationship with the phenotypic trait. The high intramodular connectivity indicates they are tightly co-expressed with other genes in the module. The “hub genes” potentially have higher gene essentiality.