ARS | Publication request: twzPEA: A Topology and Working Zone Based Pathway Enrichment Analysis Framework

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: July 13, 2010
Publication Date: N/A

Technical Abstract: Sensitive detection of involvement and adaptation of key signaling, regulatory, and metabolic pathways holds the key to deciphering molecular mechanisms such as those in the biomass-to-biofuel conversion process in yeast. Typical gene set enrichment analyses often do not use topology information in a pathway and only detects changes in mean expression levels, subject to false negatives for distributional changes without mean change. We developed a topology and working-zone based pathway enrichment analysis framework called twzPEA. The framework determines whether a general pathway is differential or conserved in comparative experimental conditions. By utilizing topology information on the pathway, specifically by decomposing interactions on pathway into homogeneous and heterogeneous components, one can increase sensitivity in detecting pathway involvement. To our knowledge, twzPEA is the first attempt on utilizing topology information of general pathways in a comparative context. Through simulation studies, we found that we have significant better power at comparing pathway involvement over a gene set enrichment analysis (GSEA) method under various noise levels. In addition, our method enables one to detect partial involvement of pathways when all interactions on pathway are not active. It is applicable to both dynamic and steady-state data. twzPEA analysis was applied to expression microarray data from two yeast strains under toxic conditions, using 62 yeast signaling and metabolic pathways from the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. Several pathways are detected as differential between the two strains including the pentose phosphate pathway. The involvement of the pentose phosphate pathway was supported by independent studies using real-time quantitative reverse transcription polymerase chain reaction (qRT-PCR).