“DompeKeys”: a set of novel substructure-based descriptors for efficient chemical space mapping, development and structural interpretation of machine learning models, and indexing of large databases

Table 7 Performances of the classification models obtained by Random Forest (RF) algorithm based on the two sets of descriptors for the three considered conjugations

Reaction	Features	Recall (BA)	MCC	Specificity (SP)
Sulfonation	DK	0.89	0.79	0.90
Sulfonation	KH	0.87	0.75	0.89
Glucuronidation	DK	0.76	0.52	0.75
Glucuronidation	KH	0.72	0.44	0.73
Reaction with glutathione	DK	0.81	0.62	0.78
Reaction with glutathione	KH	0.78	0.56	0.75
Mutagenicity	DK	0.84	0.67	0.85
Mutagenicity	KH	0.82	0.64	0.82

ISSN: 1758-2946