Peptide and protein false discovery rate - Proteome Discoverer Software - Peptide and protein false discovery rate - Proteome Discoverer Software - Proteome Discoverer Software

The false discovery rate (FDR), or the false positive rate, is a statistical value that estimates the number of false positive identifications among all identifications found by a peptide identification search. The FDR is a measure of the certainty of the identification.

You can use the application’s decoy database search feature to determine FDR.

You can use FDRs to validate MS/MS searches of large data sets.

FDRs are not effective on searches of a small number of spectra or searches against a small number of protein sequences. This is because the number of matches might be too small to give a statistically meaningful estimate.

A decoy database strategy gives both a probability value to identifiers and the percentage of false discoveries that you can expect. A good decoy database contains entries that resemble real proteins but do not contain genuine peptide sequences.

The simplest approach to obtaining such a decoy database is to reverse all protein sequences, which is the scheme that the Proteome Discoverer application uses by default. However, you can request that the search be performed against the concatenated database, when required. It is a suitable approach for enzymatic MS/MS searches.

For detailed information on how the application calculates the FDRs for PSMs, peptide groups, proteins, and protein groups, see Validation page.

For information about including the relevant nodes and their parameter settings in a workflow, see Calculating FDRs.