Ji Chen

Ji Chen, Ph.D.
07

Ph.D. Program
Quantitative Analyst
Google, Inc.

Chair

  • David States

Dissertation Title

Identification of Global and Specific Gene Expression Patterns Based on Microarray

Research Interest

Currently microarray is the most powerful tool in large-scale gene expression profiling. The ability of microarray to easily detect expression of many genes in various samples enables it to provide rich material for studies of tissue specificity; gene response evolution, etc. Also in specific targeted studies, microarray can be used to characterize genome-wide expression of different states, such as disease versus control. This application is becoming increasingly useful in pharmacogenomics research. In my thesis, I described my work in both general gene expression profiling across tissues and a targeted study. In the first part of my thesis, I introduced a novel, unbiased way of defining conservative gene response patterns based on integration of microarray data, and show that in contrast to earlier analysis methods, biological responses across multiple tissues were considered, and the responses account for a larger fraction of the variation than technical variation due to laboratory or species. 12 conservative gene expression modes (cGEM) which are connected to fundamental biological processes were found. In the second part, tissue specificity modulation of target genes of a specific transcription factor, nuclear factor kappa B (NF-kappaB), was identified based on composite microarray data from human and mouse. Bootstrap validation of clustering was used to evaluate statistical stability of tissue-associated clusters of genes. The discovery was that although NF-kappaB is a ubiquitous cellular factor, tissue variation of gene expression exists for most of its target genes, and some target genes have strong tissue specificity. In the third part, I described a specific microarray application in dental study: profiling of <italic>P. gingivalis</italic> LPS-induced gene expression in human monocytes on a genome wide basis. <italic>P. gingivalis</italic> is a Gram-negative bacterium that is recognized as one of the etiologic agents of periodontitis. In our study many genes induced by <italic>P. gingivalis </italic> LPS were found to be dependent on IKK/NF-kappaB signaling, among which were several families of transcription factors such as AP1 family proteins, nuclear receptor subfamily proteins and EGR-1 family proteins. It is highly possible that IKK/NF-kappaB may utilize these transcription factors to mediate secondary responses. Also promoter sequence analysis revealed NFkappaB binding sites in many of these induced genes, and one of them, <italic>egr-2</italic> was confirmed as a novel NF-kappaB target by CHIP assay. 

Current Placement

Google, Inc.