Table 4. Effect of competitive interactions between pairs of TFs on the accuracy of modeling ChIP data.

Data set	M2	CC(M1)	CC(M2)	CC(M1+M2)	ImprOverM1	ImprOverM2	P-value	Z-score
D_Mseq_s5_11	JIGR1	0.328	−0.277	0.368	0.040	0.091	0.02	25.90
DL_Bchip_s5	BCD	0.062	−0.116	0.193	0.131	0.077	0.01	47.80
EN_Mchip_s5_14	HTH	0.321	−0.187	0.397	0.076	0.210	0.00	17.09
EN_Mchip_s5_14	SNA	0.321	−0.129	0.362	0.041	0.233	0.03	9.23
EVE_Mseq_s14	HTH	0.174	−0.195	0.245	0.071	0.050	0.03	10.45
HB_Bchip_s9	RETN	0.204	−0.137	0.282	0.078	0.145	0.03	66.15
HB_Bchip_s9	EXD	0.204	−0.019	0.276	0.072	0.258	0.03	61.15
HB_Bseq_s5	EXD	0.374	−0.007	0.436	0.062	0.429	0.00	47.53
HB_Bseq_s5	RETN	0.374	0.023	0.420	0.046	0.398	0.02	35.84
MED_Bchip_s14	HTH	0.255	−0.178	0.360	0.105	0.182	0.00	16.81
MED_Bchip_s14	TTK-PF	0.255	−0.277	0.329	0.074	0.052	0.05	11.75
MED_Bchip_s14	SNA	0.255	−0.067	0.313	0.058	0.246	0.02	9.16
MED_Bchip_s14	HR78	0.255	−0.092	0.304	0.049	0.212	0.04	7.79
MED_Bchip_s14	HR4	0.255	−0.130	0.300	0.045	0.170	0.00	7.14
MED_Bchip_s14	HR46	0.255	−0.143	0.300	0.045	0.157	0.02	7.14
SLP1_Bchip_s5	CAD	0.214	−0.036	0.262	0.048	0.226	0.01	27.85
SLP1_Bchip_s5	EXD	0.214	0.001	0.254	0.040	0.253	0.02	23.11

STAP was used with two motifs – the primary motif and a secondary motif (“Motif 2”), but with no interaction between binding sites considered. Thus the only influence of Motif 2 on predicted occupancy of the primary TF is due to overlapping sites. Shown are all cases in which both improvements were more than 0.04, P-value≤0.05 and Z-score ≥3. The column “ImprOverM2” is the difference between CC(M1+M2) and the absolute value of CC(M2).