Consensus and Differential GO Pathway Analysis and Visualization When Multiple Transcriptomic Studies or Conditions Are Jointly Analyzed

• Activity Number: 555
• Type: Contributed
• Date/Time: Wednesday, August 12, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Chien-Wei Lin* and George C. Tseng
• Companies: University of Pittsburgh and University of Pittsburgh
• Keywords: Pathway analysis ; Meta analysis ; GO topology ; Clustering analysis

Abstract

For a single study, GO term and pathway enrichment analysis has been a widely accepted and essential part in genomic analysis. Compared to differential expression (DE) analysis for individual gene, pathway analysis usually provides broader insights for biologists to reveal the potentially underlying or relevant biological mechanism. While multiple datasets under different conditions (e.g., tissues or labs) are available, it's of interest to see whether there is any pathway enriched in one condition but not in the others. Our proposed method is based on GO pathway database, which is a well-organized database and provides the biological attributes for each gene. Due to the hierarchical structure of GO terms, different GO terms may include many overlapping genes. The corresponding GO topology was visualized through a R package named "TopGO". Biologists will know if the findings are specific to a certain condition or universal to multiple conditions. Our aim is to develop a tool that could perform differential pathway analysis under a rigorous statistical framework and also provide a visualization tool for biologists to evaluate the output in an integrated and intuitive way.