Scaffold FDR Calculations

Ever wonder why Scaffold calculates the FDR’s the way it does?

Here is a good primer on how FDR’s are calculated

Scaffold Calculates FDR’s a bit differently

Here is the inside scoop From BrianĀ Searl from proteome software

“When you calculate FDR=2*D/(T+D) you assume that decoys are still in
the results list and the FDR you’re interested in is for all proteins
identified, including decoys, which is why the denominator is T+D. We
presume that you know (and can filter out) which proteins are decoys
(they’re tagged in pink) and you’re really interested in the FDR of
only the remaining proteins in your list (the T target proteins). You
could formalize that as FDR=I/(C+I) where C is the count of correct
proteins and I is the count of incorrect proteins, where implicitly
C+I=T, or the total number of targets. Scaffold make the assumption
that D=I, or that you have an equal number of decoy proteins as
incorrect proteins from your original search, giving you FDR=D/T.”


Posted in Proteomics Blog

Leave a Reply

Your email address will not be published. Required fields are marked *