Analyzing omics data of biological samples

In the projects I participated in the VirtualLiver Network, chip array experiments were performed, mostly on the Affymetrix platform to obtain mRNA transcript profiles. As the final aim of VirtualLiver are multi-scale modeling, these experiments served as a resource to identify genes whose expression is affect by the various tretments of interest. These treatments are

• tested on cell culture of primary cell cultures:
  • cytokines occurring during liver regeneration. HGF, IL6, TGFβ, TNFα tested
  • cytokines occurring during liver inflammation. TGFβ
  • agonists of the nuclear receptors PXR, CAR, PPARα which are triggered in liver intoxication and also in liver development
• tested on mouse liver tissue material:
  • fractioned periportal vs. pericentral hepatocytes
  • genetically modified mouse with altered Hedgehog signaling
• different stages of obesity, steatosis, and stetohepatitis obtained from human liver biopsies

The analyses are:

• ANOVA and t-test to assess the singnificance of the changes
• ModeScore analysis to assert which metabolic functions are affected
• Show the most changed genes, for the assumption that the amplitude of change corresponds roughly to the importance of the process
• Show the most changed genes in each of the gene classes
• Assert synergy, attenuation, divergence, and dominance in the co-treatment of several factors