Skip to main content

Table 1 Performance comparisons in terms of numbers of differentially expressed genes.

From: A statistical framework for consolidating "sibling" probe sets for Affymetrix GeneChip data

 

Mult-test Algorithm

Unequal Variance

Equal Variance

Per-Gene

Bonferroni (.05)

39

45

 

FDR-BH (.05)

313

434

 

FDR-BY (.05)

63

84

 

RawP cut-off (6.5e-05)

124

151

Per-ProbeSet

Bonferroni (.05)

2

6

 

FDR-BH (.05)

6

59

 

FDR-BY (.05)

1

4

 

RawP cut-off (6.5e-05)

6

59

customCDF-UniGene

Bonferroni (.05)

3

18

 

FDR-BH (.05)

16

87

 

FDR-BY (.05)

1

8

 

RawP cut-off (6.5e-05)

18

40

customCDF-ensEMBLgene

Bonferroni (.05)

6

10

 

FDR-BH (.05)

10

103

 

FDR-BY (.05)

1

10

 

RawP cut-off (6.5e-05)

18

59

  1. Comparison of the per-gene approach, the per-probeset approach, the UniGene custom CDF approach, and the ensEMBL gene custom CDF approach in terms of screening differentially expressed genes between wild type and Nrl knockout.
Back to article page

BMC Genomics

ISSN: 1471-2164

Contact us