Position specific variation in the rate of evolution in transcription factor binding sites

Table 2 Evolution of motifs with known consensus, but binding sites identified by MEME

Cluster	Factor	Consensus identified by MEME	Motif subs.	Bg subs.	Corr.	p-value
Protein folding +	Hsf1p	TTTTC TAGAA AGTTC	0.14	0.68	-0.42	0.060
Glycolysis +	Gcr1p	AAATAGAGGAAG CCCA	0.23	0.63	-0.80	<0.001 **
Nitrogen +	Gln3p/ Dal80p	TCTTAT CA	0.39	0.74	-0.78	0.010 *
Gluco-neogenesis +	Sip4p (?) CSRE	CC GTTTGTCCG	0.33	0.57	-0.84	<0.001 **
G1 phase +	Mbp1p Swi6	TTACGCGT TTT	0.22	0.64	-0.67	0.011 *
Respiration +	Hap2/3/4p	TGATTGG TCCA	0.20	0.67	-0.53	0.048 *
Methionine +	Cbf1p	ATGTCACGTG	0.13	0.75	-0.49	0.078
Proteasome +	Rpn4p	ATTTTGCCACC G	0.20	0.73	-0.75	0.002 *
M/G1 transition +	Swi5p/ Ace2p	AACCAGC A	0.26	0.61	-0.57	0.074
Repressed in Stress ++	(?) PAC	ATGCGATGAG CTGAG	0.24	0.71	-0.69	0.006 *
leu/ilv bio-synthesis++	Leu3p	GCCG TTTCCGG	0.31	0.70	-0.54	0.044 *
Phosphate +++	Pho4p	CCCACGTG CG	0.29	0.65	-0.74	0.005 *
119 positions in all computationally identified motifs					-0.60	4e-14 **

Here binding sites are identified by running the MEME program [3] on genes that clustered with targets in micro-array gene expression data. Expected consensus sequences (from [20, 40] or [7]) are underlined. 'Motif subs.' and 'bg subs.' are the substitutions per site in the binding sites and the promoters in which they are found respectively. 'Corr.' and 'p-value' are the Spearman's rank correlation coefficient and the associated p-value between the rate of evolution at each position and the information content at each position. * Indicates significance at a per factor error rate of < 0.05. ** Indicates significance after Bonferoni correction for a global error rate < 0.05, assuming 50 tests were done in total. (?) indicates uncertainty as to the identity of the binding protein. + indicates clusters taken from hierarchical clustering [40] of yeast data from the Stanford Microarray database [42], ++ indicates clusters taken from hierarchical clustering of 300 genetic perturbations [43] and +++ indicates clusters taken from hierarchical clustering of 64 control experiments [43]

ISSN: 2730-7182