BMC Bioinformatics

Table 2 F1, MCC, TPr, TNr, PRC using the full set of features ( α ) or only the features selected ( β ) in Table 1

From: Transient protein-protein interface prediction: datasets, features, algorithms, and the RAD-T predictor

	LR						AT
	Full set			NI1			Full set			NI1
	α	β		α	β		α	β		α	β
MCC	0.1247	0.1304	*	0.1684	0.188	***	0.1376	0.1418		0.1859	0.2012	***
TPr	0.6462	0.6497		0.6449	0.6597	*	0.7064	0.7615	***	0.7958	0.8047
TNr	0.5386	0.5435	**	0.5435	0.5519		0.4915	0.4428		0.4049	0.4149	***
PRC	0.2082	0.2108	*	0.3445	0.3543	***	0.2092	0.2053		0.3328	0.3383	***
F1	0.2946	0.2971		0.4359	0.447	**	0.3019	0.304		0.4562	0.4635	***
	BN						FT
	Full set			NI1			Full set			NI1
	α	β		α	β		α	β		α	β
MCC	0.1356	0.1412		0.188	0.2043	**	0.1319	0.1408	**	0.1607	0.1981	***
TPr	0.7794	0.8553	***	0.7866	0.842	***	0.704	0.7545	***	0.6931	0.821	***
TNr	0.4143	0.3341		0.4197	0.3749		0.4881	0.4504		0.485	0.3905
PRC	0.2008	.1976		0.3358	0.3341		0.2059	0.2055		0.335	0.3343
F1	0.3005	0.3007		0.4574	0.4654	**	0.2983	0.3034	***	0.4385	0.4628	***
	RT
	Full set			NI1
	α	β		α	β
MCC	0.1059	0.1196	***	0.128	0.1769	***
TPr	0.6333	0.6583	***	0.6246	0.6986	***
TNr	0.5227	0.5173		0.5221	0.502
PRC	0.1999	0.2041	**	0.327	0.343	***
F1	0.2843	0.2922	***	0.417	0.4461	***

Algorithm abbreviations in Table 1. The algorithms were applied to the full protein set (Full Set) and the NI1 subset. Statistical significance was calculated using using 1-sided Wilcoxon’s signed rank test. (* P < 0.05; ** P < 0.01; *** P < 0.001).

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com