Skip to main content
. 2020 May 4;21:173. doi: 10.1186/s12859-020-3485-y

Table 2.

Results achieved with the three methods for estimating variant FDRs on simulated data, with the FDR control level at 1%

n iCombined FDR iSeparate FDR iTransfer FDR
Ave.#I.D.s Est.error(%) Ave.#I.D.s Est.error(%) Ave.#I.D.s Est.error(%)
(false/all) Mean SD (false/all) Mean SD (false/all) Mean SD
1 10.22/10.70 -94.36 5.13 0.21/0.55 -16.95 34.64 0.00/0.25 0.00 0.02
5 10.76/13.09 -80.95 8.84 0.30/2.05 -11.09 20.57 0.00/1.24 0.01 0.07
10 10.09/14.87 -66.43 9.50 0.60/4.17 -12.92 11.63 0.00/2.50 0.04 0.14
20 10.90/20.57 -51.64 9.08 0.72/ 8.07 -8.86 6.31 0.00/5.30 0.08 0.19
50 10.53/34.44 -29.25 6.87 0.73/19.07 -3.60 2.57 0.00/13.28 0.20 0.27
100 10.67/58.64 -17.09 4.54 0.73/38.27 -1.43 1.25 0.00/27.40 0.42 0.35

Note: n, the number of variant mass spectra; Ave.#I.D.s, average number of false/all identifications of variant peptides from the target database at 1% estimated FDR; Est.error, the difference between the estimated FDR and the real FDR; Mean and SD, mean and standard deviation of the Est.error as percentage