Skip to main content
. 2019 Oct 31;10:2464. doi: 10.3389/fmicb.2019.02464

TABLE 9.

Comparing PointFinder and PhyResSE prediction performance.

Drug Res Sus PointFinder
PhyResSE
p-value
Spec. Sens. MCC Spec. Sens. MCC
RMP 14 77 0.961 1.000 0.890 0.935 1.000 0.830 0.751
INH 29 62 0.952 0.897 0.848 0.968 0.931 0.899 0.262
STM 37 54 0.981 0.649 0.693 0.981 0.838 0.843 0.047
EMB 14 77 0.961 0.857 0.796 0.974 0.857 0.831 0.395
PZA 8! 83! 0.964 0.750 0.677 0.964 0.625! 0.589 0.666

! Indicates that data varied from the previously published results, since the number of resistance and susceptible samples found in the published Supplementary Data did not correlate with amount assessed after rerunning the data through PheResSE. Res, number of resistant strains determined by drug susceptibility testing; Sus, number of susceptible strains determined by drug susceptibility testing; Spec, specificity; Sens, sensitivity; MCC, Mathew Correlation Coefficient; RMP, Rifampicin; INH, Isoniazid; STM, Streptomycin; EMB, Ethambutol; PZA, Pyrazinamide. It was tested weather PhyResSE was performing significantly better using bootstrapping, and the p-value is given, and the significance level was set at 0.05 indicated with .