Protein complexes identification based on go attributed network embedding

BMC Bioinformatics

Table 3 Performance comparison based on six evaluation metrics on the five yeast data

Datasets	Methods	#predicted complexes	#matched complexes	Precision	Recall	F-score	Acc
DIP	COACH	570	263	0.450	0.620	0.521	0.243
	CMC	179	108	0.603	0.394	0.477	0.219
	MCODE	59	32	0.542	0.118	0.194	0.149
	ClusterOne	341	133	0.390	0.343	0.365	0.227
	MCL	451	69	0.153	0.172	0.162	0.190
	PEWCC	666	413	0.620	0.469	0.534	0.230
	GANE	324	202	0.623	0.550	0.584	0.254
Krogan-core	COACH	348	206	0.592	0.460	0.518	0.217
	CMC	128	86	0.672	0.304	0.419	0.206
	MCODE	71	52	0.732	0.198	0.311	0.176
	ClusterOne	522	190	0.364	0.464	0.408	0.273
	MCL	376	126	0.335	0.414	0.371	0.262
	PEWCC	630	425	0.675	0.406	0.507	0.214
	GANE	208	161	0.774	0.436	0.558	0.229
Krogan14k	COACH	570	263	0.461	0.465	0.463	0.217
	CMC	396	187	0.472	0.440	0.455	0.210
	MCODE	49	30	0.612	0.112	0.189	0.152
	ClusterOne	225	105	0.467	0.302	0.366	0.222
	MCL	445	133	0.299	0.323	0.311	0.233
	PEWCC	934	500	0.535	0.418	0.470	0.217
	GANE	247	169	0.684	0.442	0.537	0.234
Biogrid	COACH	1507	469	0.311	0.657	0.422	0.276
	CMC	1503	236	0.157	0.553	0.245	0.265
	MCODE	58	16	0.276	0.043	0.075	0.181
	ClusterOne	476	187	0.393	0.497	0.439	0.316
	MCL	338	77	0.228	0.219	0.223	0.249
	PEWCC	2781	1044	0.375	0.677	0.483	0.288
	GANE	637	347	0.545	0.664	0.599	0.310
Collins	COACH	251	188	0.749	0.522	0.615	0.280
	CMC	153	104	0.680	0.390	0.496	0.255
	MCODE	111	94	0.847	0.400	0.540	0.254
	ClusterOne	195	143	0.733	0.511	0.602	0.290
	MCL	183	134	0.732	0.506	0.598	0.286
	PEWCC	570	477	0.837	0.426	0.564	0.252
	GANE	199	163	0.819	0.491	0.615	0.293

Both F-score and Acc are overall evaluation metrics, so the highest values of F-score and Acc are set in bold for each dataset

ISSN: 1471-2105