Predicting protein complex in protein interaction network - a supervised learning based method

BMC Systems Biology

Table 7 Performance comparison between the Equal Weight (EW) and Regression models (RM)

merg_ thred	model	Num	Ncp	P	Ncb	R	F	Sn	PPV	Acc
0.1	EW	167	94	0.5629	182	0.2725	0.3672	0.6075	0.6550	0.6308
	RM	212	144	0.6792	274	0.4102	0.5115	0.3473	0.7849	0.5221
0.2	EW	227	109	0.4802	217	0.3249	0.3875	0.6092	0.6635	0.6358
	RM	286	184	0.6434	317	0.4746	0.5462	0.4054	0.7749	0.5605
0.3	EW	289	127	0.4394	234	0.3503	0.3898	0.6190	0.6329	0.6259
	RM	333	206	0.6186	327	0.4895	0.5465	0.4242	0.7713	0.5720
0.4	EW	336	151	0.4494	254	0.3802	0.4119	0.6256	0.6393	0.6324
	RM	388	242	0.6237	339	0.5075	0.5596	0.4466	0.7557	0.5809
0.5	EW	405	188	0.4642	269	0.4027	0.4313	0.6229	0.6387	0.6307
	RM	491	308	0.6273	350	0.5240	0.5710	0.4649	0.7317	0.5832
0.6	EW	475	221	0.4653	279	0.4177	0.4402	0.6197	0.6328	0.6262
	RM	624	393	0.6298	353	0.5284	0.5747	0.4779	0.7335	0.5920
0.7	EW	568	261	0.4595	285	0.4266	0.4425	0.6201	0.6353	0.6277
	RM	755	493	0.6530	359	0.5374	0.5896	0.4908	0.7370	0.6014
0.8	EW	674	298	0.4421	288	0.4311	0.4366	0.6203	0.6362	0.6282
	RM	880	570	0.6477	363	0.5434	0.5910	0.4938	0.7363	0.6030
0.9	EW	819	348	0.4249	288	0.4311	0.4280	0.6195	0.6395	0.6294
	RM	945	584	0.6180	363	0.5434	0.5783	0.4942	0.7350	0.6027

Ncp denotes the number of the correct predictions that match at least a true complex and Ncb denotes the number of the true complexes that match at least one predicted complex.

ISSN: 1752-0509