TABLE 2.

Correlation of phenotype and model predictions of SVM and SCM models.

Antimicrobial agent	Number of isolates classified by	SVM		SCM
	phenotype (R, S)	Number of isolates predicted to be TP and TN (TP, TN)	Correlation of phenotype to the model prediction	Number of isolates predicted to be TP and TN (TP, TN)	Correlation of phenotype to the model prediction
Tetracycline	(58, 38)	(55, 38)	p < 2.2e−16	(50, 38)	p < 2.2e−16
Ampicillin	(19, 77)	(19, 77)	p < 2.2e−16	(19, 77)	p < 2.2e−16
Sulfisoxazole	(46, 50)	(46, 50)	p < 2.2e−16	(46, 50)	p < 2.2e−16
Trimethoprim	(16, 80)	(16, 80)	p < 2.3e−16	(16, 80)	p < 2.3e−16
Enrofloxacin	(6, 90)	(6, 90)	p < 1.1e−09	(6, 90)	p < 1.1e−09

R resistant, S sensitive. Correlation between phenotype and model predictions were calculated using Fisher’s exact test in python.