Predicting RNA-binding residues from evolutionary information and sequence conservation

BMC Genomics

Table 3 Prediction performance breakdown in terms of the categories of RNA using the training dataset, RB147

Predictor	RNA	TP	FP	TN	FN	Sensitivity	Specificity	Precision	Accuracy	MCC	F-score	F_0.5-score
ProteRNA_SVM	rRNA	2060	537	5814	1856	52.60%	91.54%	79.32%	76.69%	0.4933	0.6326	0.7201
mRNA	27	16	1606	229	10.55%	99.01%	62.79%	86.95%	0.2193	0.1806	0.3154
tRNA	234	171	11000	996	19.02%	98.47%	57.78%	90.59%	0.2942	0.2862	0.4105
others	109	93	6930	646	14.44%	98.68%	53.96%	90.50%	0.2441	0.2278	0.3487
Total	2430	823	25344	3727	39.47%	96.86%	74.70%	85.92%	0.4741	0.5165	0.6338
ProteRNA_WildSpan	rRNA	554	412	5939	3362	14.15%	93.51%	57.35%	63.24%	0.1274	0.2270	0.3560
mRNA	67	121	1501	189	26.17%	92.54%	35.64%	83.49%	0.2139	0.3018	0.3323
tRNA	50	173	10998	1180	4.07%	98.45%	22.42%	89.09%	0.0566	0.0688	0.1178
others	85	272	6751	670	11.26%	96.13%	23.81%	87.89%	0.1045	0.1529	0.1947
Total	756	978	25189	5401	12.28%	96.26%	43.60%	80.27%	0.1489	0.1916	0.2887
ProteRNA	rRNA	2256	878	5473	1660	57.61%	86.18%	71.98%	75.28%	0.4618	0.6400	0.6856
mRNA	89	138	1484	167	34.77%	91.49%	39.21%	83.76%	0.2764	0.3685	0.3823
tRNA	238	304	10867	992	19.35%	97.28%	43.91%	89.55%	0.2431	0.2686	0.3502
others	177	366	6657	578	23.44%	94.79%	32.60%	87.86%	0.2118	0.2727	0.3024
Total	2760	1686	24481	3397	44.83%	93.56%	62.08%	84.28%	0.4376	0.5206	0.5764

ISSN: 1471-2164