DRaW: prediction of COVID-19 antivirals by deep learning—an objection on using matrix factorization

BMC Bioinformatics

Table 3 Comparison of DRaW with the other methods on COVID-19 datasets

Datasets	Methods	Recall	Specificity	Precision	F1 score	AUC-ROC	AUPR
DS1	IRNMF	0.750	0.614	0.182	0.2927	0.706	0.2927
	VDA-KLMF	0.892	0.544	0.300	0.367	0.939	0.763
	VDA-RWLRLS	0.562	0.838	0.141	0.225	0.885	–
	DRaW	0.642	0.836	0.651	0.620	0.822	0.589
DS2	IRNMF	0.801	0.728	0.220	0.345	0.816	0.2933
	VDA-KLMF	0.826	0.531	0.208	0.283	0.857	0.377
	VDA-RWLRLS	0.513	0.826	0.007	0.123	0.835	-
	DRaW	0.513	0.778	0.441	0.463	0.865	0.458
DS3	IRNMF	0.741	0.771	0.174	0.2820	0.809	0.222
	VDA-KLMF	0.863	0.522	0.163	0.233	0.866	0.391
	VDA-RWLRLS	0.519	0.843	0.067	0.118	0.862	–
	DRaW	0.538	0.847	0.576	0.550	0.887	0.558

ISSN: 1471-2105