Table 2.

Source datasets of the combined training resource

Dataset	Plates	Fields of view	Treatment Type	Treatments	Queries	Ground truth	Purpose	Reference
LINCS	136	23,094	Compounds	129	N/A	N/A	Training only	Way et al.7
BBBC043	16	7,992	Gene ORFs	50	N/A	N/A	Training only	Caicedo et al.10
BBBC036	55	122,022	Compounds	1550	1,365	Mechanism of action	Training and evaluation	Bray et al.38
BBBC022	20	66,558	Compounds	995	849	Mechanism of action	Training and evaluation	Gustafsdottir et al.37
BBBC037	5	17,254	Gene ORFs	205	204	Genetic pathway	Training and evaluation	Rohban et al.36