A FC-GSEA Approach to Identify Significant Gene-Sets Using Microarray Gene Expression Data - Advances in Computational Science and Engineering - page 121

SNR Fisher's criterion
Fig. 6. An example of ES located in a negative region for original GSEA but located in a posi-
tive region for FC-GSEA
[CASE 3] When ES in original GSEA is taken from a positive or negative SNR
region and ES in FC-GSEA is taken from a negative FC region
This case reflects the situation that the maximum deviation ES point in FC-GSEA
occurs in a negative region. Since the FC values are always positive, the negative
quantity of ES means that there are no matching genes with a given gene-set. Thus, in
such a case, we take the maximum deviation from zero in a positive region as ES in
FC-GSEA for a gene-set, and significant genes are taken from the left side of the
maximum deviation ES point. On the other hand, in original GSEA, significant genes
are taken from the left or right side of the maximum deviation ES point, depending on
whether the ES point is located either in a positive region or in a negative region.
SNR Fisher's criterion
Fig. 7. An example of ES located in a negative region for FC-GSEA
Thus, by using FC-based gene ranking for the gene-set enrichment analysis, it is
possible to select both highly up-regulated genes and highly down-regulated genes at
the same time as the significant genes for a gene-set. On the other hand, recall that the
SNR-based gene ranking chooses either only the highly up-regulated genes or only
the highly down-regulated genes as significant ones for a gene-set.