Skip to main content

Table 2 Performance of GSEA and SAM-GS on Test 2. Proportions of randomly generated non-null gene sets that are identified by each method to be associated with the phenotype (p-value ≤ 0.05) in a mouse-microarray study.

From: Improving gene set analysis of microarray data by SAM-GS

Pearson correlation of genes in the gene set with the phenotype

Methods

Set Size

  

10

30

50

100

Half of genes with |r| ≥ .4, the other half with |r| < .4

GSEA

1%

3%

0%

1%

 

SAM-GS

93%

100%

100%

100%

Half of genes with |r| ≥ .5 the other half with |r| < .5

GSEA

3%

4%

3%

1%

 

SAM-GS

100%

100%

100%

100%

Half of genes with |r| ≥ .6, the other half with |r| < .6

GSEA

6%

7%

7%

18%

 

SAM-GS

100%

100%

100%

100%

Half of genes with |r| ≥ .7 the other half with |r| < .7

GSEA

12%

18%

31%

66%

 

SAM-GS

100%

100%

100%

100%

Half of genes with |r| ≥ .8, the other half with |r| < .8

GSEA

20%

64%

88%

100%

 

SAM-GS

100%

100%

100%

100%

Half of genes with |r| ≥ .9, the other half with |r| ≥ .9

GSEA

69%

100%

100%

100%

 

SAM-GS

100%

100%

100%

100%