Feature request: Detailed reporting of identified fragment ions

[StefanTenzer]

Dear Fragpipe Team, thanks for your amazing work! We are currently trying to extract information at fragment ion level from FragPipe results. Ideally, we would need a report file that provides addition information at the fragment ion level for each identified PSM (in the psm.tsv) , i.e. observed product_MZ, theoretical_product_MZ, Intensity, Annotation. This file will likely be relatively large in size, but ideal for e.g. visualization of the identified spectra.

Best regards,

Stefan Tenzer

[fcyu]

Hi Stefan,

Thank you very much for your suggestion. If you want to visualize the annotated spectra, you can open the FP-PDV window by clicking "Open visualization window" in the "Run" tab.

Best,

Fengchao

[StefanTenzer]

Dear Fengchao, thank you for the swift reply - we would ideally need this info in a machine readable format (i.e. tsv), visualization was only an example of possible use cases. We e.g. would like to compare Fragpipe results and results from PEAKS on fragment ion level, and do some more in-depth analyses. For that, we tried to map back from library.tsv -> psm.tsv -> spectrum, but this is not very efficient and exact. I.e. library.tsv seems to integrate all fragment information from all spectra of the same peptide, so the information in which specific spectrum the fragment was identified is lost.

Best wishes, Stefan

[fcyu]

I have implemented the feature to report the matched fragments. Let MSFragger generate the tsv report will have those two new columns. If you want to try the pre-released version, please send an email to yufeATumich.edu (replace AT with @).

Currently, there are some limitations

1. The reported fragments are not exactly the one used in the scoring because MSFragger does not record such information (to be fast). I let it "re-match" the fragments when generating the report.
2. The fragments are deisotoped by default, so the reported matches only have singly charged. Disabling the deisotoping in the database search will have doubly charged fragments being matched, but the sensitivity is sub-optimal.
3. It only supports closed search now, but I will extend it to open and mass-offset search in the future.

Best,

Fengchao