Inferring circRNA-drug sensitivity associations via dual hierarchical attention networks and multiple kernel fusion

Table 1 Performance comparison based on five-fold cross-validation

Method	F1-Score	Accuracy	Recall	Specificity	Precision	AUC	AUPR
Dataset: data271
DHANMKF(our)	0.8500	0.8520	0.8648	0.8636	0.8618	0.9178	0.9262
MNGACDA [17]	0.8472	0.8424	0.8723	0.8155	0.8247	0.9139	0.9209
GATECDA [15]	0.8224	0.8186	0.8404	0.7966	0.8054	0.8873	0.8915
MKGCN [35]	0.8230	0.8208	0.8221	0.8193	0.8350	0.8768	0.8984
MMGCN [39]	0.8190	0.8183	0.8231	0.8135	0.8156	0.8766	0.8664
MINIMDA [37]	0.7988	0.7901	0.8331	0.7472	0.7684	0.8562	0.8534
LAGCN [38]	0.7900	0.7786	0.8338	0.7233	0.7516	0.8505	0.8478
GANLDA [40]	0.7936	0.7822	0.8384	0.7259	0.7542	0.8517	0.8468
Dataset: data251
DHANMKF(our)	0.8597	0.8588	0.8655	0.8521	0.8552	0.9263	0.9286
MNGACDA [17]	0.8444	0.8429	0.8528	0.8330	0.8362	0.9136	0.9211
MKGCN [35]	0.8253	0.8247	0.8285	0.8208	0.8222	0.8941	0.9090
MMGCN [39]	0.8065	0.8120	0.7836	0.8404	0.8308	0.8842	0.9049
GATECDA [15]	0.8336	0.8287	0.8225	0.8354	0.8451	0.8802	0.9034
MINIMDA [37]	0.8053	0.7950	0.8476	0.7424	0.7670	0.8575	0.8502
LAGCN [38]	0.7880	0.7857	0.7966	0.7748	0.7796	0.8540	0.8615
GANLDA [40]	0.7883	0.7956	0.8089	0.7823	0.7880	0.8539	0.8551

ISSN: 1471-2164