Table A3. Logo plots of clusters of odd palindromic motifs from the E. coli phylogenetic footprints (Bayes ratio: 6-8)

Cluster ID	TF^a	# motifs^b	# sites^c	Strength^d	Genes^e	Sequence logo^f
14	?	4	14	7.92	frdA, yhbL, yheS, ydgR
25	?	3	13	7.63	atpI, yfeH, yfeU
1	?	7	33	6.89	fadD, yhhW, b2342, acpD, hflB, ydhB/ydhC, lipA
56	?	2	10	6.78	flgA/flgB, pcnB
21	?	3	12	6.76	gapA/yeaA, tufA, thrS
38	?	2	11	6.75	pepA, pssR
12	?	4	18	6.72	b0947/ycbY, yhdG, icdA, yfjB
6	?	5	25	6.69	ssb, rnhA/dnaQ, lpp, map/rpsB, artP
7	?	4	17	6.53	glyA, tnaA, ytfE, cysP
52	ompR	2	8	6.48	b2343, b2532/suhB
16	?	3	15	6.39	rpsL, tig, rplM
31	?	3	12	6.33	lig, yfiM, cysW
32	nagC	2	11	6.11	nagE/nagB, ybfM

^a The transcription factors that have been experimentally determined to bind to E. coli sites in the motifs listed in column 6.

^b Number of motifs (genes) in the cluster.

^c Total number of binding sites (known and putative) in the cluster.

^d This is defined as the log-Bayes ratio normalized by the number of motifs in the cluster. Bayes ratio is the ratio of the probability of the data belonging to one cluster to the probability of the data as separate motifs.

^e The names of the E. coli genes corresponding to the motifs in each cluster (i.e., the motifs were identified in the promoter regions of these genes; many of these genes are known to be the first gene of an operon, but the downstream genes of those operons are not listed). Gene names in black indicate that the E. coli site has been experimentally confirmed as a binding site for the TF in column 2, unless more than one type of TF is listed, in which case the gene names are color coded to match the correct TF; gene names in red have not been experimentally confirmed as a binding site. The names of divergently transcribed genes are shown separated by a “/”, (e.g. ydhB/ydhC).

^f Created by Schneider & Stephens’software²⁰ from http://www.lecb.ncifcrf.gov/~toms/logoprograms.html