Table 4.

Summary of the performance of each algorithm using CGH microarray data for the MG1655 sequenced strain. The sensitivity, specificity, and M-score generated from each of the cut-off algorithms from the CGH data were summarized for comparison.

Algorithm	M-Score	Sensitivity	Specificity
Naive Cut-off (0.25)	99.12	99.07	99.31
Mixture Model (Bimodal)	99.07	99.23	98.63
MKD	98.99	98.93	99.18
Porwollik	98.75	98.37	99.86
GENCOM	98.45	97.99	99.79
GACK EPP = 0	98.43	97.97	99.79

* Institute of Food Research method (GENCOM)

** Genotyping Analysis by Charlie Kim method (GACK)

*** Minimum Kernel Density method (MKD)