Figure 1

Schematic representation of the analysis principles and differences between eGSA and other array analysis approaches. Both individual gene analysis (IGA) and Gene Set Analysis (GSA) approaches are based on the gene-phenotype correlation. In GSA, the correlation value is mapped to GSs, and then summarized by the scoring functions such as averaged t-statistic or K-S test [5]. The significance of each score is then estimated by statistic tests [1]. In IGA, differentially expressed genes (DEGs) are selected based on the correlation threshold, and GS significance is estimated by enrichment statistics. In eGSA, the gene expression levels of test samples (test 1–6) are converted to expression regulatory events (REs) by comparing with the reference sample pool (reference 1–6). The sum of RE frequency is then used for GS scoring and estimation of the significant levels.