Pipeline Methods for Global Analysis of Meta Expression Data: A study of the Herpesviridae family

Abstract

Developing new tools for computational analysis of related expression data sets led us to significant findings in human cytomegalovirus (HCMV) and Kaposi's sarcoma- associated herpesvirus (KSHV) suggesting unique means by which the viruses interface with the MYC gene and its associated pathways. Our methods assayed multiple expression data sets from three different herpesviruses, in various cell types, and flagged gene sets and communities that were most affected by virus treatment. Community labeling by what we call functional annotation, further streamlined an analysis pipeline underscored by gene set enrichment analysis (GSEA) and revealed proof of concept findings as well as new ones. Ultimately, we find that KSHV infection affects the MYC pathway by downregulating the expression of MYC target genes. In HCMV, we find and show that while expression of most MYC targets is up during infection, expression of the MYC gene itself is down. We offer hypotheses to explain this phenomenon as well as analysis of other specific findings as they pertain to the computational assay. This study generates an improvement to pipelines seeking to analyze expression data and provides a means by which to allow concurrent study to take place over multiple studies of similar or different nature. This enables more robust study going forward, and a method for re- analyzing previously collected data more specifically and thoroughly.