Skip to main content
. 2011 Jun 22;6(6):e19749. doi: 10.1371/journal.pone.0019749

Table 2. Diagnostic parameters of Minim typing for CC8 and subCCs within CC8.

CC/subCC STsa MelTsb Sensitivity Specificity PPV
Rawc Enhancedd Raw Enhanced Raw Enhanced
8 (all) 185 22 0.973 0.965 0.976 0.983 0.863 0.906
8–8 110 10 0.891 0.935 0.992 0.998 0.899 0.972
8–239 44 7 0.911 0.948 0.992 0.997 0.788 0.932
8–72 14 1 0.717 0.733 0.996 0.996 0.630 0.688
8–247 11 0e ND ND ND ND ND ND
8–770 6 2 0.830 0.897 0.997 0.997 0.560 0.625

The calculations assumed equal abundance of all STs. ND = not determined.

a

Number of STs belonging to the CC/subCC.

b

Number of MelTs diagnostic for the CC/subCC. Each MelT defines either a single ST belonging to the CC/subCC of interest, or a group of STs of which at least 50% are of the CC/subCC of interest.

c

The CC and subCC assignments for STs in the Excel key were used without modification.

d

Singletons that are DLVs of the founder of the CC/subCC were classified as true positives, and false positive and false negative STs that are SLVs for more than one CC/subCC founder were assigned to the CC/subCC that maximizes the diagnostic parameters. Additional singletons that are founder DLVs and do not correspond to MelT diagnostic for the CC/subCC of interest were identified as new false negatives.

e

There are no MelTs diagnostic for subCC 8–247 according to the criterion used, primarily because CC7 and subCC 8–247 are not discriminated by Minim typing.