Table 2.

Evaluation of gene-level circular RNA prediction based on the 5-fold cross-validation

Method	Accuracy	Precision	Sensitivity	Specificity	F1-score	MCC	AUC
SVM	0.7121 ± 0.0560	0.8346 ± 0.0596	0.5836 ± 0.1958	0.8534 ± 0.1067	0.6652 ± 0.1212	0.4662 ± 0.0719	0.7185 ± 0.0504
RF	0.7324 ± 0.0060	0.7035 ± 0.0097	0.8499 ± 0.0041	0.6018 ± 0.0131	0.7697 ± 0.0050	0.4688 ± 0.0105	0.7259 ± 0.0058
Att-CNN	0.7246 ± 0.0051	0.7555 ± 0.0275	0.8302 ± 0.0454	0.5519 ± 0.0624	0.7898 ± 0.0088	0.4006 ± 0.0071	0.6910 ± 0.0091
Att-RNN	0.7300 ± 0.0076	0.7565 ± 0.0246	0.8340 ± 0.0407	0.5639 ± 0.0546	0.7923 ± 0.0086	0.4164 ± 0.0153	0.6989 ± 0.0099
nRC	0.7290 ± 0.0086	0.7378 ± 0.0329	0.7592 ± 0.0594	0.6961 ± 0.0663	0.7461 ± 0.0154	0.4591 ± 0.0190	0.8013 ± 0.0076
PredcircRNA	0.6188 ± 0.0033	0.6594 ± 0.0085	0.5732 ± 0.0085	0.6696 ± 0.0117	0.6132 ± 0.0049	0.2433 ± 0.0077	0.6085 ± 0.0314
circDeep	0.8387 ± 0.0066	0.8778 ± 0.0135	0.8063 ± 0.0087	0.8749 ± 0.0133	0.8404 ± 0.0073	0.6806 ± 0.0138	0.7522 ± 0.0107
DeepCirCode	0.8629 ± 0.0215	0.8940 ± 0.0268	0.8424 ± 0.0606	0.8860 ± 0.0374	0.8659 ± 0.0265	0.7296 ± 0.0377	0.8642 ± 0.0198
JEDI	0.9659 ± 0.0059	0.9670 ± 0.0075	0.9685 ± 0.0166	0.9630 ± 0.0095	0.9676 ± 0.0057	0.9318 ± 0.0119	0.9657 ± 0.0054

Note: We report the mean and standard deviation for each metric.