Table 1.

In silico models for the development of novel antibacterial compounds.

No.	Ntotal	Nantibiotics	Number of variables	Techniquea)	Overall accuracyb) (%)	Ref.
1	111	60	7	LDA	93.8/91.5**	(Garcia-Domenech and de Julian-Ortiz, 1998)
				ANN	89.0/97.9**
2	664	249	62	ANN	94.8**	(Tomas-Vert et al., 2000)
3	59	24	17	LDA	85.0/84.0***	(Mishra et al., 2001)
4	661	249	6	LDA	92.6/93.6*	(Cronin et al., 2002)
				BLR	94.7/94.3*
5	664	249	3	LDA	90.1**	(Aptula et al., 2003)
				BLR	92.1**
6	351	213	7	LDA	91.0/89.0***	(Molina et al., 2004)
7	433	217	6	LDA	85.7/87.5**	(Murcia-Soler et al., 2004)
			62	ANN	98.7/91.4**
8	667	363	7	LDA	92.9/94.0**	(Gonzalez-Diaz et al., 2005)
9	657	249	34	ANN	92.9**/100.0***	(Cherkasov, 2005)
10	2,030	1,006	8	LDAc)	90.4/89.3**/93.1***	(Marrero-Ponce et al., 2005)
11	4,346	520	62	kNN	95.0/95.0*/84.4***	(Karakoc et al., 2006)
12	611	230	36	SVC	100.0*/100.0**/98.1***	(Yang et al., 2009)
				kNN	97.7**/96.1***
				DT	98.6*/92.3**/91.0***
13	7,517	2,066	21	kNNc)	99.2*/81.8**/78.3***	(Wang et al., 2014)
14	2,230	1,051	3	LDAc)	85.6/87.2**/86.2***	(Castillo-Garit et al., 2015)
15	3,500	628	4	ISE	94.6/72.0***	(Masalha et al., 2018)

^a)LDA, linear discriminant analysis; ANN, artificial neural network; BLR, binary logistic regression; kNN, k-nearest neighbors; MLR, multiple linear regression; SVC, support vector classification; DT, decision tree; ISE, iterative stochastic elimination. ^b)*Cross-validation; **internal test set; ***external test set. ^c)Models that demonstrated the highest quality with external test set