. 2009 May 25;37(Web Server issue):W606–W611. doi: 10.1093/nar/gkp288

Table 1.

MGIP is more sensitive and faster than other commonly used methods

	TP	FN	Sn (%)	Speed (s)^a
MGIP versus STARS^b
MGIP	660	18	97.4	63 ± 0.58
STARS	653	35	94.9	323 ± 30
MGIP versus SeqMan method^c
MGIP	323	6	98.2	75 ± 2
SeqMan	319	8	97.6	1520 ± 173

TP: true positives; FN: false negatives; Sn: sensitivity.

^aSpeed is shown as an average per trace file set (84 traces in the STARS comparison, 96 traces in the SeqMan comparison), plus or minus standard deviation. The speed tests were performed over a 1 Gigabyte per second network connection and therefore the upload time was negligible. However, the upload time from a slower connection will understandably increase the time to process a set of trace files. Approximate times for uploading a set of traces is given in Supplementary Table 2.

^bFor the MGIP versus STARS comparison, 17 sets of MLST data were tested which were composed of trace files over 691 strain/loci. The speed test was performed on three randomly selected sets, composed of 126 strain/loci.

^cFor the MGIP versus SeqMan method comparison, 10 sets of fetA were tested in the SeqMan comparison, totaling 331 loci. The speed test was performed on three randomly selected sets, composed of 103 strain/loci.