Bacterial strain tracking when used to identify strain sharing in twin infant pairs. Test results for previously published tools (MIDAS, StrainPhlAn and inStrain) were obtained from the literature 28 . Classification as the same strain of conspecific strains residing in twin infants by each of the tools was considered as a true-positive classification, while classification of strains as the same in unrelated infants was considered as a false-positive classification. The AUC is an estimator of the method performance, in terms of the ratio of true-positive rate and false-positive rate at different ANI (MIDAS, inStrain and StrainPhlAn) or APSS (SynTracker) values. n represents the total number of strain-pair comparisons performed by each tool.
