Table 4.
Replacing regulatory coupling strength by response to transcription factor deletion.
| Transcription Factor | Number of genes in each group | GO category | ||||
| B- | B+/D+ | B+/D- | Not available | B+/D+ | B+/D- | |
| Dig1 | 5881 | 1 | 30 | 115 | 32 | 16 |
| Gcn4 | 5928 | 16 | 63 | 60 | 19 | 17 |
| Hir2 | 5941 | 2 | 19 | 105 | 24 | 5 |
| Mbp1 | 5647 | 2 | 103 | 315 | 22 | 12 |
| Swi4 | 5535 | 19 | 106 | 407 | 9 | 6 |
| Swi5 | 5368 | 10 | 82 | 607 | 2 | 0 |
| Yap1 | 5968 | 7 | 37 | 55 | 0 | 0 |
We analyzed the performance of a scheme in which our TF-gene coupling factor was replaced by the change in mRNA abundance in response to transcription factor deletion as a predictor of true targets. Significant binding (B+) was defined as before, while a significant response to deletion (D+) required P < 10-2. The number of significantly over-represented GO categories is listed for both the B+/D+ and the B+/D- gene groups. The results indicate that TF deletion data is less useful than our coupling factor for distinguishing functional target genes from non-functional ones.