PDRLGB: precise DNA-binding residue prediction using a light gradient boosting machine

BMC Bioinformatics

Table 3 Performance comparison of various prediction methods on PDNA-62 with 5-fold cross-validation

Methods	ACC	SN	SP	ST	PRE	F1	MCC	AUC	P-value
Dps-pred	0.791	0.403	0.818	0.611	0.279	0.330	0.191	-	-
Dbs-pssm	0.664	0.682	0.660	0.671	0.210	0.376	0.249	-	-
BindN	0.703	0.694	0.705	0.700	0.291	0.410	0.297	0.752	-
Dp-bind	0.781	0.792	0.772	0.782	0.378	0.512	0.490	-	-
BindN-RF	0.782	0.781	0.782	0.782	0.385	0.516	0.436	0.861	-
BindN+	0.790	0.773	0.793	0.783	0.395	0.523	0.443	0.859	-
PreDNA	0.794	0.768	0.797	0.783	0.398	0.524	0.424	-	-
EL _PSSM-RT	0.808	0.854	0.801	0.826	0.428	0.569	0.507	0.901	**
PDRLGB	0.815	0.863	0.806	0.835	0.438	0.581	0.523	0.912	1.79 ×10⁻⁵

ISSN: 1471-2105