TABLE 3.

Summary of TFs identified using different approaches

	Effect					No effect					Not determined
		Maintained site					Maintained site					Maintained site
Approach	No site	Conserved	Preserved	Relocated	Total	No site	Conserved	Preserved	Relocated	Total	No site	Conserved	Preserved	Relocated	Total
Consensus	16	3	1	1	21	91	20	2	20	133	133	9	4	25	171
ChIP cons.	5	2	0	0	7	5	2	1	1	9	4	1	0	0	5
Variation	16	4	5	18	43	—	—	—	—	—	—	—	—	—	—
ChIP var.	1	0	0	1	2	9	0	1	9	19	7	2	2	7	18
Total	38	9	6	20	73	105	22	4	30	161	144	12	6	32	194

Approach: Consensus, site identified by the presence of a perfect match to a reported consensus; ChIP cons, site identified by the presence of a perfect match to a reported consensus as well as by ChIP–chip data; variation, TF identified by the R-SGA assay with an alteration from the known consensus; ChIP var., site identified by ChIP–chip data with an alteration from the known consensus; effect, effect was observed in either the R-SGA assay or reported data (four TFs spanning five sites were added to ChIP cons.); no effect, no effect was found when TF was deleted in the R-SGA assay; not determined, gene not present or not grown in the array. Proteins that bind as a complex were referred to as a single TF. These are: Gcr1-2, Hap2-5, Ino2-4, Rtg1-3, and Swi4-6. Maintained sites, the presence of TF-binding sites was examined in S. cerevisiae, S. bayanus, and S. paradoxus and characterized as follows: conserved, identical sequence in all strains; preserved, sequence was not identical, but consensus was present; relocated, sequence was found at an adjacent site, up to 200 bp; no site, TF-binding site was not found in the three examined strains. Results calculated from Table S1. Additional information is given in Fig. S1 and Table S1.