search

Nesvilab / FragPipe

A cross-platform proteomics data analysis suite
http://fragpipe.nesvilab.org
Other
209 stars 38 forks source link

Spectral Library Generation: PeptideProphet failing #1839

Closed tryhard124 closed 1 month ago

tryhard124 commented 1 month ago

Hi all,

I'm having a similar proplem to that deatiled in this thread so I thought I'd open up a new issue. I double-checked and I didn't have any problems with my FASTA file, so I'm not sure why PeptideProphet is failing. I'm attaching my log and fasta file (in my workflow i have it as a fasta.fas file but converted to .txt to attach it here) in case y'all might be able to help. This is a spectral library for a single protein using DDA data generated on TF Orbitrap Ascend.

log_2024-10-23_17-43-33.txt

2024-10-23-decoys-Martin_P37840.txt

Cheers, Martín

anesvi commented 1 month ago

Your database has only 42 peptides. Is it just a single protein ? Tools like Percolator or PeptideProphet are not applicable when too few PSMs. You can see if percolator works but I doubt. You need to search a bigger database or not use those tools and not apply FDR filters

tryhard124 commented 1 month ago

Yes it's just a single protein. I'm trying to build up a PRM method for this single protein and its PTMs. I was trying to build a spectral library but since the list of peptides is short, do I just manually selecting precursors w/ rt's from the data since the list is so short? I was under the impression that it's still needed for data processing. As you predicted, percolator failed when I orginally tried to process this data on PD so I used Fixed Value PSM validator.

anesvi commented 1 month ago

I am not sure. Yes you can look at MSFragger output. You can also try to search a bigger database. Often there are a many contaminant or copurifying proteins, which would give enough PSMs for modeling. Or use Skyline instead of FragPipe, they have search engines incorporated, and must have a good way to build PRM assays