Journal of Cheminformatics

Table 1 The 17 target class-specific datasets used in this study

From: LigTMap: ligand and structure-based target identification and activity prediction for small molecular compounds

Target class	Core set			Benchmark set^a
Target class	Total	Training	Validation	Benchmark set^a
Human target
Kinase	2008	1608	400	18
Transferase	559	448	111	–
Beta-secretase	309	248	61	19
Hydrolase	1196	957	239	–
Anticoagulant	264	212	52	–
Carbonic anhydrase	354	285	69	16
Ligase	89	71	18	5
Bromodomain	167	133	34	19
Isomerase	110	91	19	–
Estrogen	76	61	15	–
Peroxisome	16	13	3	–
Diabetes	99	81	18	–
Non-human target
HIV	524	419	105	10
Tuberculosis	232	186	46	11
HCV	159	127	32	–
Influenza	99	81	18	–
Helicobacter pylori	52	41	11	–
Total	6313	5062	1251	98

^aFor the benchmark set, where no new suitable data were found in literature, the entries are marked as “–.” Sources of benchmark data are: Kinase [46], Ligase [47], BRD [45], CA [48], beta-secretase [49], HIV [42], and TB [50]

Back to article page

ISSN: 1758-2946

Contact us

Submission enquiries: journalsubmissions@springernature.com