Predicting tissue specific cis-regulatory modules in the human genome using pairs of co-occurring motifs

BMC Bioinformatics

Table 1 Mining for heart-specific CRMs - input data

n	CNEs/Locus	CNEs	Controls
0	9.6	1480	7400
10	10.8	1667	8335
20	13.1	2019	10095
30	15.3	2358	11790
40	17.5	2692	13460
60	20.9	3219	16095
80	23.7	3644	18220
100	25.7	3959	19795
All	31.9	4911	24555

The mixed sets consisted of CNEs found in the vicinity of 154 heart-specific genes. Nine sets were constructed. All mixed sets contained CNEs within 10 kbp upstream and 10 kbp downstream of the transcription start sites (TSSs). We expanded the mixed sets by including n CNEs that are closest to the TSSs, if they were not already included, n = 0, 10, 20, 30, 40, 60, 80, 100, ∞. A control set included five times as many sequences as its mixed set counterpart. We controlled for sequence length while assembling the control sets.

ISSN: 1471-2105