Exploiting bounded signal flow for graph orientation based on cause–effect pairs

Table 2 Thresholded network parameters

threshold	n	m	p	n _t	p _t	n*	m _v	m _e	q	q _v	q'	$q_{v}^{'}$	n _c	m _c	k
0.000000	5385	39921	14393	799	2014	750	2014	59	7	7	3	3	115	292	17
0.154420	4530	35041	11522	747	2203	705	2203	298	27	27	20	20	475	1632	40
0.371369	4254	32135	10740	796	2443	749	2443	275	47	47	35	35	528	2424	46
0.573290	3871	27128	9445	777	2225	704	2225	268	32	32	13	13	140	311	32
0.573313	2546	8977	5279	638	2311	477	2310	208	252	252	151	151	561	2394	68
0.830093	2206	7136	4346	643	2206	449	2206	192	304	304	193	193	727	4017	83
0.886308	1407	3646	1607	441	787	260	785	45	106	106	88	88	311	1876	75
0.943001	1135	3069	920	361	464	195	463	32	57	57	42	42	179	801	44
0.954421	1039	2504	843	350	489	175	461	45	85	73	71	61	215	3001	81
0.957338	895	2060	681	304	405	119	375	39	64	54	58	50	240	3092	89
0.965986	874	2018	666	299	477	103	411	165	90	78	85	75	358	12284	110
0.984753	668	1676	312	206	163	95	162	20	7	7	6	6	55	222	15
0.989212	581	1322	188	192	167	69	161	86	24	24	24	24	141	1088	32
0.989233	307	681	71	121	70	32	66	36	21	21	11	11	52	219	7
0.990409	294	666	28	114	27	26	26	21	2	2	2	2	9	8	2

Parameters for the largest connected component of the protein interaction network assembled by Nir Yosef [25] with different thresholds for the edge probability. The uneven gaps in the sizes of the instances are because many edges have identical weights.

ISSN: 1748-7188