Identification of protein functions using a machine-learning approach based on sequence-derived properties

Table 8 Detailed results of the random forest method without feature selection (RF_FF)

Protein class	Sensitivity	Specificity	F-measure	MCC
Transport	87.58	95.89	0.84	0.91
Transcription	96.87	92.4	0.89	0.95
Translation	56.25	100	0.74	0.72
Gluconate utilisation	85.71	100	0.92	0.92
Amino acid biosynthesis	96.89	95.65	0.92	0.95
Fatty acid metabolism	71.6	100	0.82	0.84
Acetylcholine receptor inhibitor	100	100	1.00	1.00
G-protein coupled receptor	95.44	99.11	0.95	0.97
Guanine nucleotide-releasing factor	85.71	100	0.92	0.92
Fibre protein	83.33	100	0.91	0.91
Transmembrane	95.07	99.3	0.95	0.97

ISSN: 1477-5956