Table 1. Evaluation of target protein prediction using common data (a) and merged data (b).

a. common data
similarity search strategy	CMap - Biological control					LINCS - Level 3
	signature : all genes		signature : L1000 genes		signature : L1000 genes
	AUC	AUPR	AUC	AUPR	AUC	AUPR
same cell line-matching	0.8288	0.1213	0.8310	0.1279	0.8203	0.1145
different cell line-matching	0.8178	0.1027	0.8283	0.1161	0.8099	0.0922
similarity search strategy	CMap - Mean centering				LINCS - Level 4
	signature : all genes		signature : L1000 genes		signature : L1000 genes
	AUC	AUPR	AUC	AUPR	AUC		AUPR
same cell line-matching	0.8261	0.1228	0.8257	0.1248	0.8145		0.1115
different cell line-matching	0.8176	0.1036	0.8224	0.1179	0.8077		0.0992
b. merged data
similarity search strategy	CMap - Biological control				LINCS - Level 3
	signature : all genes		signature : L1000 genes		signature : L1000 genes
	AUC	AUPR	AUC	AUPR	AUC		AUPR
same cell line-matching	0.5648	0.0232	0.5650	0.0241	0.8608		0.0985
different cell line-matching	0.5640	0.0206	0.5649	0.0228	0.8195		0.0505
similarity search strategy	CMap - Mean centering				LINCS - Level 4
	signature : all genes		signature : L1000 genes		signature : L1000 genes
	AUC	AUPR	AUC	AUPR	AUC		AUPR
same cell line-matching	0.5643	0.0235	0.5643	0.0231	0.8517		0.0819
different cell line-matching	0.5641	0.0206	0.5645	0.0215	0.8245		0.0537