Protein-RNA interface residue prediction using machine learning: an assessment of the state of the art

BMC Bioinformatics

Table 9 Top Six Methods on Structure Data using Residue-Based Evaluation

	PSSMSeq	Smo PSSMSeq	PSSMSeq	Smo PSSMSeq	Smo PSSMStr	Smo PSSMStr
	RBFK	RBFK	LK	LK	RBFK	LK
RB106Str	0.81 (1)	0.79 (3)	0.78 (4.5)	0.77 (6)	0.80 (2)	0.78 (4.5)
RB144Str	0.81 (1)	0.79 (4)	0.79 (4)	0.77 (6)	0.80 (2)	0.79 (4)
RB198Str	0.80 (1)	0.79 (2.5)	0.78 (4.5)	0.77 (6)	0.79 (2.5)	0.78 (4.5)
Average	0.80 (1)	0.79 (3.2)	0.78 (4.3)	0.77 (6)	0.80 (2.2)	0.78 (4.3)

Comparison of AUC (averaged over five folds) of the top six methods on structure data using residue-based evaluation. Based on average rank, the best method across all three datasets is the SVM classifier that uses the RBF kernel and PSSMSeq as input. (NB - Naïve Bayes, SVM - Support Vector Machine, LK - Linear Kernel, RBFK - Radial Basis Function Kernel) The rank of each classifier is shown in parentheses.

ISSN: 1471-2105