Table 2. Predicting phenotypic resistance using genetics.

	TP	FP	TN	FN	VME	ME	PPV	NPV
INH	5,493	142	5,622	224	0.039	0.025	0.961	0.975
RIF	4,535	435	6,669	107	0.023	0.061	0.977	0.939
EMB	1,919	513	6,702	111	0.055	0.071	0.945	0.929
LEV	1,689	255	8,104	184	0.098	0.031	0.902	0.969
MXF	1,358	504	9,022	160	0.105	0.053	0.895	0.947
AMI	632	84	10,117	163	0.205	0.008	0.795	0.992
KAN	735	124	9,043	197	0.211	0.014	0.789	0.986
ETH	971	114	9,183	511	0.345	0.012	0.655	0.988

Statistics on how much resistance can be explained in a dataset enriched for rare resistance mutations using a standard resistance catalogue that predates the CRyPTIC project. TP, the number of phenotypically resistant samples that are correctly identified as resistant (“true positives”); FP, the number of phenotypically susceptible samples that are falsely identified as resistant (“false positives”); TN, the number of phenotypically susceptible samples that are correctly identified as susceptible (“true negatives”); FN, the number of phenotypically resistant samples that are incorrectly identified as susceptible (“false negative”); VME, very major error rate (false-negative rate), 0–1; ME, major error rate (false-positive rate), 0–1; PPV, positive predictive value, 0–1; NPV, negative predictive value, 0–1.

AMI, amikacin; EMB, ethambutol; ETH, ethionamide; INH, isoniazid; KAN, kanamycin; LEV, levofloxacin; MXF, moxifloxacin; RIF, rifampicin.