Functional Based Analysis and Visualization of Gene Expression Data from Hepatocytes Grown on Different Substrates

Abstract

The ability to grow a fully functional Hepatocyte in vitro would be a prodigious breakthrough for modern day drug testing and regenerative medicine. However there are many factors affecting Hepatocyte functionality in vitro that need to be studied before this is possible, like the effect of the underlying substrate that the Hepatocyte is grown on. To better understand this interaction, gene expression profiles of Hepatocytes were measured on four different substrates: Corning(r) CellBIND(r), Collagen, Corning Matrigel(r) and Locust Bean Gum (S906). Here, we present data analysis methods that enable data reduction, identification of differentially expressed genes, and grouping of genes with similar intensities by implementing data clustering. The optimum number of clusters in a given data set is estimated using the gap statistic method. We further employ Cytoscape to help visualize the data and incorporate GeneMania to help identify potentially interesting patterns across substrates. One of the interesting findings from our analysis suggests up-regulation of genes involved in vasculature development on all four substrates. Such findings can be used to design new experiments that can provide further insight on the interactions between hepatocytes and their substrates and help understand the mechanism behind responses of hepatocytes on these substrates.