Table 2

Our result is more biologically meaningful than Hu et al.’s and Jiang et al.’s results
Test (with FDR = 0.05) Our result vs. Hu et al.’s result Our result vs. Jiang et al.’s result
Functional enrichment Our result has an equal or higher enrichment score than Hu et al.’s result in 83% (93/112) of the cases. Our result has an equal or higher enrichment score than Jiang et al.’s result in 83% (29/35) of the cases.
Expression coherence Of the 112 TFs, 86% (96/112) show significantly higher expression coherence in our result, compared with only 2% (2/112) in Hu et al.’s result. Of the 35 TFs, 49% (17/35) show significantly higher expression coherence in our result, compared with only 31% (11/35) in Jiang et al.’s result.
The prevalence of protein-protein interactions Of the 112 TFs, 82% (92/112) are enriched for membership to a protein-protein interaction module in our result, compared with only 38% (43/112) in Hu et al.’s result. Of the 35 TFs, 91% (32/35) are enriched for membership to a protein-protein interaction module in our result, compared with only 31% (11/35) in Jiang et al.’s result.

To show that our result is more biologically meaningful than Hu et al.’s and Jiang et al.’s results, we assessed the functional enrichment, the expression coherence, and the prevalence of protein-protein interactions. Our result outperforms their results across all measures. Note that the comparison between our result and Jiang et al.’s result was performed on 35 TFs since only the TF knockout targets of these 35 TFs were reported in Jiang et al.’s study.

Yang and Wu

Yang and Wu BMC Systems Biology 2012 6:102   doi:10.1186/1752-0509-6-102

Open Data