Hybrid coexpression link similarity graph clustering for mining biological modules from multiple gene expression datasets

Table 1 Topological analysis of edge clusters for varying β

		Edge clusters					0.5-sup		0.75-sup		1.0-sup
β	\|E _S\|	\|M\|	$\bar{V}$	$\bar{E}$	$\bar{RE}$	$\bar{ρ}$	$sup^{¯}$	\|M ^′\|(%)	$sup^{¯}$	\|M ^′\|(%)	$sup^{¯}$	\|M ^′\|(%)
0	2273510	421	48.5	77.2	0.74	0.16	6.8	337(80)	1.9	46(11)	0.3	0(0)
0.1	1945908	470	44.3	69.1	0.74	0.16	6.9	382(81)	2	57(12)	0.3	0(0)
0.2	873741	1559	15.8	19	0.58	0.25	9.1	1546(99)	4	578(37)	1.1	11(1)
0.3	226832	1777	10.5	10.4	0.41	0.28	10.3	1777(100)	5.9	1380(78)	2.4	209(12)
0.4	55897	896	8.5	8.2	0.36	0.31	11.5	896(100)	7.8	880(98)	4.1	383(43)
0.5	15779	276	9.5	9.4	0.32	0.3	11.7	276(100)	8.9	275(100)	5.4	162(59)
0.6	3864	147	7.7	7.4	0.33	0.33	12.6	147(100)	10	147(100)	6.9	127(86)
0.7	625	25	6.2	6.8	0.45	0.45	15.1	25(100)	12	25(100)	9.1	24(96)
0.8	625	4	5.2	6.5	0.66	0.66	19	4(100)	14.2	4(100)	10.5	4(100)
0.9	20	1	4	6	1	1	23	1(100)	22	1(100)	20	1(100)

An α value of 0.5 is used and only edge clusters with at least 4 edges are retained.

ISSN: 1756-0381