Ancient Pbx-Hox signatures define hundreds of vertebrate developmental enhancers

BMC Genomics

Table 2 Frequency of KR motifs in CNEs at different gene loci

GENE	# KR motifs in test set	Length of CNE seq for locus (kb)	#hits per kb	# hits in control set (mean)	standard deviation	z-score	p-value
ZNF503	36	27.781	1.30	3.18	1.76	18.62	0.00E+00
TSHZ3	30	23.323	1.29	3.09	1.77	15.23	0.00E+00
IRX5	27	37.059	0.73	5.39	2.33	9.29	0.00E+00
IRX2	21	23.981	0.88	3.10	1.80	9.95	0.00E+00
TSHZ1	16	10.351	1.55	1.63	1.32	10.93	0.00E+00
PBX3	16	17.886	0.89	1.89	1.35	10.44	0.00E+00
HOXD9	16	17.77	0.90	2.19	1.44	9.59	0.00E+00
NR2F2	16	18.99	0.84	2.52	1.59	8.49	0.00E+00
NR2F1	16	25.655	0.62	3.72	1.84	6.67	2.53E-11
MEIS2	16	24.553	0.65	3.42	1.91	6.59	4.49E-11
ZFHX1B	13	23.275	0.56	3.13	1.72	5.73	9.86E-09
SALL3	12	11.405	1.05	1.43	1.21	8.76	0.00E+00
FOXP1	12	15.857	0.76	1.73	1.24	8.25	2.22E-16
MAF	11	7.334	1.50	1.15	1.10	8.95	0.00E+00
NKX6-1	10	6.853	1.46	0.82	0.92	9.94	0.00E+00

Details are shown for the 15 gene loci from the CONDOR CNE set with the highest number of Pbx-Hox KR motifs in their CNEs, showing enrichment relative to shuffled CNE sets (Methods and Additional File 6). For each gene locus, the number of Pbx-Hox KR motifs in the associated CNEs is given. The number of Pbx-Hox KR motifs per kb of CNE sequence for each locus (column 4) is calculated by dividing the number of Pbx-Hox KR motifs in the CNEs of that locus (column 2) by the total combined length of the CNEs in that locus (column 2). Control sets were generated by zero order Markov shuffling of CNEs at each locus in 1000 randomisations (Methods). Some gene loci also contain other genes besides the one after which they are named, for instance the IRX5 locus contains Irx3, Irx5 and Irx6.

ISSN: 1471-2164