Meta-learning-based Inductive logistic matrix completion for prediction of kinase inhibitors

Table 2 Performance comparison of ILMC & MTDNN on head & tail kinase of LTKinaseDB (global evaluation model)

		AUC	AUPR	BA	RECALL	PRECISION	F1
ILMC (MACCS + CTD)	Head	0.8386	0.7919	0.7571	0.7590	0.7062	0.7317
ILMC (MACCS + CTD)	Tail	0.7204	0.4509	0.6621	0.6423	0.3995	0.4926
ILMC (ECFP + ProtVec)	Head	0.8549	0.8193	0.7876	0.7832	0.7448	0.7635
ILMC (ECFP + ProtVec)	Tail	0.7049	0.4646	0.6595	0.6675	0.4102	0.5277
MTDNN [20]	Head	0.8753	0.8401	0.7945	0.7868	0.7552	0.7706
MTDNN [20]	Tail	0.7034	0.4448	0.6477	0.6639	0.3724	0.4772
MolTrans [29]	Head	0.8453	0.8041	0.7643	0.7679	0.6774	0.7488
MolTrans [29]	Tail	0.7253	0.4442	0.6805	0.6908	0.3620	0.5377

ISSN: 1758-2946