Prosecutor: parameter-free inference of gene function for prokaryotes using DNA microarray data, genomic context and multiple gene annotation sources

Table 2 Extending transcriptional modules of E. coli

transcriptional module	gene	prosecutor rank	motif rank	motif sequence in the intergenic region of either the gene or its operon	literature reference
ArgR Amino acid biosynthesis: Arginine. AUC 0.92	artJ	21	8	TGCATAACATTGCG	[56]
	aroP	58	39	TGATTTTTAATTCA	[57]
	artI	131	50	TGCATAATTATTCT	[56]
	hisL	16	4	TGAATAAACATTCA	putative
	pyrL	32	61	TGACTTTTAATTCA	putative
	metH	36	76	TGAATTTTTATTAA	putative
	ydcS	43	63	TGAATAAATTTTCT	putative
	stpA	132	21	TGCATTTTTATTCA	putative
	hisG	141	8	TGAATAAACATTCA	putative
	hisJ	144	27	TGCATTGAAATGCA	putative
	hisC	145	13	TGAATAAACATTCA	putative
	hisA	147	14	TGAATAAACATTCA	putative
	potF	162	46	TGCATAAAAATTTG	putative
CysB Amino acid biosynthesis: Cysteine AUC 0.91	sbp	12	0	CGCAAGTTATAGCCAATCTTTTTTTATTCTT	[48, 58]
	nlpA	36	17	CAGACTTTATATTCCACTTTTATTCCTTTTT	[48]
	mmuP	28	40	AACGCGGTATAACAAACCTTCTTTGGATGTT	putative
Fur iron regulatory gene AUC 0.84	yncD	74	32	GGGAATGGTAATCATTATT	[44]
	ybaN	37	5	GAAAATGATAATTGTTATG	putative
	folE	101	29	GGCAATTACAATAATTATC	putative
LexA major regulator of DNA repair AUC 0.87	yebG	0	21	CTGTATAAAATCACAG	[59, 60]
	dinI	2	6	CTGTATAAATAACCAG	[61, 62]
	dinB	6	51	CTGTATACTTTACCAG	[63]
	dinD	19	0	CTGTATATAAATACAG	[45]
	yjiW	39	20	CTGATGATATATACAG	[45]
	ybfE	120	31	CTGATTAAAAACCCAG	[45]
	sbmC	125	4	CTGTATATAAAAACAG	[64]
MetJ Amino acid biosynthesis: Methionine AUC 0.88	ybdH	10	6	AGACGTTTAGATGTCT	[65]
	ybdL	106	0	AGACATCTAAACGTCT	[65]
	ycbK	198	17	AGTCATCTTGACGTCT	[65]
	mmuP	14	15	GGATGTTTAGATGTCC	putative

Transcriptional module member predictions identified by Prosecutor applied to well-known transcriptional modules for E. coli. Gene: the gene for which a validated functional prediction with a transcriptional module (column three) was found. Rank: represents the position of a gene in the sorted list with p-values. These p-values describe the functional prediction significance for every individual gene with a specific functional category. More significant p-values are matched with lower ranks. AUC: describes the association efficiency for a transcriptional module with respect to its own members. In addition to the rank information provided by Prosecutor, supplemental motif information is provided. This data is obtained by applying a position specific scoring matrix (PSSM) to the upstream sequences of all genes in the genome. The PSSM is derived from aggregating all known consensus target sequences (DNA regulatory binding sites). The additional motif information allows users to concentrate on genes that exhibit coexpression with a transcriptional module as well as possessing a predicted consensus sequence. This additional evidence contributes to the confidence in assigning a gene to a particular transcriptional module. Motif rank: based on the results for a PSSM when matched to every upstream sequence in the genome. For example, based on the PSSM of the regulator LexA, the upstream region of gene dinD contains the best ranking motif (rank 1).

ISSN: 1471-2164