Christie Hunter


How does the scoring and FDR analysis work for SWATH acquisition data processing? For proteomics

Similar extraction processes are used in both the SWATH acquisition microapp in PeakView software on the desktop and the Extractor app in OneOmics suite.

After the fragment ions have been selected for each peptide, the extracted ion chromatograms are generated and the sets of XICs generated make up the peptide groups that are scored for each peptide. Scores are computed using both chromatographic and spectra components. The chromatographic components to the score include consistency of peak apex RT, consistency of peak width at half max, consistency of XIC peak area with library fragment ion pattern, etc.

For the spectral components to the score, first a spectrum is generated for each peak group by taking the spectra at the peak apex and subtracting the background spectra on each side of the peak. This double background subtracted spectrum is then scored using spectral components including mass accuracy of mono-isotopic peak, confirmation of mono-isotopic and presence of contiguous ion series matching library spectrum.

Once all the peptides are scored, a false discovery rate analysis is performed in a very similar way to the standard target – decoy analysis that is done in proteomics. For every target sequence, a pseudo-reverse sequence (decoy) is generated (pseudo because we reverse the sequence but keep the C-term the same). Both the target and the decoy sequences are scored as above. They are then ranked based on this score. Then the FDR is computed by 2*# of decoys in the list/total (or position of the target sequence in the list). The FDR can then be assigned to each peptide and used when exporting results such that the user has control over the level of error in their results.


Join the discussion

Comment below on this article and our team will answer your questions.


Submit a Comment