miTAR: a hybrid deep learning-based approach for predicting miRNA targets

BMC Bioinformatics

Table 4 Performance comparison between miTAR2 and miRAW on an independent test dataset

Model:dataset	Accuracy	Sensitivity	Specificity	F-score	PPV^b	NPV^b	Brier score
miRAW: miRAWRaw^a	0.913	0.931	0.363	0.954	NA	NA	NA
miTAR2: miRAWIn	0.966	0.957	0.976	0.967	0.977	0.956	0.028

NA represents the value is not reported in the corresponding study
^amiRAWRaw presents the dataset used in the miRAW study. The best performance values were selected from the reports of the miRAW study under different conditions
^bPPV represents positive predictive value; NPV represents negative predictive value

ISSN: 1471-2105