sdm-help-list > Error in FDR calculation
Showing 1-3 of 3 posts
Sep 8, 2020 12:09 PM | Mikhail Schelkunov
Error in FDR calculation
There is an error in FDR calculation at https://www.sdmproject.com/utilities/?sh... .
If you paste numbers:
0.001052588
0.002887613
0.004249
0.2431061
0.4909836
0.589325
The q-value 0.008498 will appear in the output twice.
Similarly, if you paste numbers:
6.80E-08
8.26E-08
2.86E-07
4.45E-07
1.09E-06
0.001052588
0.002887613
0.004249
0.2431061
0.4909836
0.589325
The q-value 4.543e-07 will appear twice.
If you paste numbers:
0.001052588
0.002887613
0.004249
0.2431061
0.4909836
0.589325
The q-value 0.008498 will appear in the output twice.
Similarly, if you paste numbers:
6.80E-08
8.26E-08
2.86E-07
4.45E-07
1.09E-06
0.001052588
0.002887613
0.004249
0.2431061
0.4909836
0.589325
The q-value 4.543e-07 will appear twice.
Sep 8, 2020 02:09 PM | Joaquim Radua
RE: Error in FDR calculation
Dear Mikhail
R returns the same FDRs:
[1] 0.006315528 0.008498000 0.008498000 0.364659150 0.589180320 0.589325000
[1] 4.543000e-07 4.543000e-07 1.048667e-06 1.223750e-06 2.398000e-06 1.929745e-03 4.537678e-03 5.842375e-03 2.971297e-01 5.400820e-01 5.893250e-01
With best wishes,
Joaquim
R returns the same FDRs:
[1] 0.006315528 0.008498000 0.008498000 0.364659150 0.589180320 0.589325000
[1] 4.543000e-07 4.543000e-07 1.048667e-06 1.223750e-06 2.398000e-06 1.929745e-03 4.537678e-03 5.842375e-03 2.971297e-01 5.400820e-01 5.893250e-01
With best wishes,
Joaquim
Sep 9, 2020 07:09 AM | Mikhail Schelkunov
RE: Error in FDR calculation
Thank you, that's my fault. I compared the results on this site
with the results on https://tools.carbocation.com/FDR and observed a difference. It appears that the bug is in their
algorithm, not yours.