Table 3.

TF interaction connectance comparison between networks before and after using our approach.

		Genie3_1,500			Genie3_2,000			Genie3_5,000			Reference network
		Better	Worse	Equal	Better	Worse	Equal	Better	Worse	Equal	#edges	#nodes
Curated PPI		0.373	0.333	0.294	0.346	0.222	0.433	0.381	0.255	0.364	796	271
ARCHS4	corr_0.25	0.620	0.093	0.286	0.511	0.065	0.424	0.546	0.099	0.355	30,915	344
	corr_0.45	0.603	0.111	0.287	0.492	0.084	0.425	0.520	0.125	0.355	16,584	326
	corr_0.65	0.569	0.143	0.287	0.462	0.112	0.425	0.482	0.161	0.356	6,912	256
	corr_0.85	0.492	0.187	0.321	0.412	0.145	0.443	0.430	0.194	0.377	353	95
STRING	combined_0.5	0.352	0.355	0.293	0.306	0.264	0.43	0.341	0.298	0.361	1065	260
	combined_0.8	0.297	0.392	0.311	0.336	0.225	0.439	0.362	0.265	0.373	241	150
	textmining_0.4	0.344	0.366	0.29	0.285	0.287	0.428	0.317	0.323	0.36	1351	265
	textmining_0.6	0.299	0.392	0.309	0.291	0.274	0.435	0.332	0.302	0.366	502	196
	textmining_08	0.191	0.461	0.348	0.275	0.273	0.452	0.315	0.305	0.38	223	139
	experimental_05	0.184	0.468	0.348	0.3	0.241	0.459	0.333	0.278	0.388	120	117
	experimental_07	0.1	0.48	0.419	0.233	0.27	0.498	0.255	0.315	0.43	49	65
	experimental_09	0.047	0.428	0.525	0.148	0.281	0.571	0.165	0.336	0.499	15	26
	database_04	0.228	0.434	0.337	0.334	0.208	0.459	0.358	0.249	0.392	69	53

This table shows the percentage of genes with greater connectance in the interaction network for all its TFs in all interaction networks employed to test how using GENIE3 to filter the networks improved the three GRN based on distance TF assignment (1.5, 2, and 5 kb at maximum between the TFBS and its target gene).