Table 1. AUC results of our method, CENTDIST, CEAS, and CORE_TF on three ChIP-seq datasets with size more than 20000 peak points.

	Our method	CENTDIST	CORE_TF promBGa			CORE_TF randBGb			CEASc
			200	400	1000	200	400	1000	200	400	1000
ESRRB	0.8892	0.7869	0.6373	0.6627	0.6065	0.5359	0.5451	0.6183	0.6203	0.6072	0.6111
E2F1	0.8828	0.8761	0.8202	0.7966	0.7758	0.8076	0.7862	0.7303	0.5789	0.5625	0.5746
TCFCP	0.8437	0.9072	0.6889	0.6719	0.5386	0.6627	0.6484	0.6641	0.6333	0.6144	0.6105

^aCORE_TF promBG under promoter background uses enriched regions with size 200, 400 and 1000.

^bCORE_TF randBG under random genome background uses enriched regions with size 200, 400 and 1000.

^cCEAS uses enriched regions with size 200, 400 and 1000.