Is there a way to obtain the unfiltered peptide list

[PernilleTJensen]

I ran quantms with a higher false discovery rate (FDR) threshold in order to increase the number of peptide quantified. Unexpectedly, increasing the FDR from, for example, 0.10 to 0.25 resulted in fewer reported peptides.

Code used:
nextflow run bigbio/quantms
--input XXXXX.sdrf.tsv
--database XXXX.fasta
--outdir XXXXX
--add_decoys
--decoy_string DECOY_
--protein_level_fdr_cutoff 0.1
--psm_level_fdr_cutoff 0.1
-profile singularity

Question:
Is there a way to obtain the unfiltered peptide list (i.e., prior to FDR filtering), or to better understand why increasing the FDR threshold leads to fewer peptide identifications?

[ypriverol]

Can you explain a bit more about how you run the workflows? It shouldn't report fewer peptides if PSM threshold is relaxed. Can you provide the parameters?

[PernilleTJensen]

I ran the pipeline on a server using the following command:
nextflow run bigbio/quantms
--input XXXXX.sdrf.tsv
--database XXXX.fasta
--outdir XXXXX
--add_decoys
--decoy_string DECOY_
--protein_level_fdr_cutoff 0.1
--psm_level_fdr_cutoff 0.1
-profile singularity

I tested different values for --protein_level_fdr_cutoff and --psm_level_fdr_cutoff (e.g. 0.10 and 0.25), but increasing the thresholds did not increase the number of reported peptides in the output file: quant_tables/XXXX.sdrf_openms_design_msstats_in.csv

Is there another parameter or filtering step in the pipeline that I may have overlooked which affects peptide-level output or FDR filtering?

[jpfeuffer]

Hi!

There is also run_fdr_cutoff since quantms does a two step filtering (first per run, then global).
You can find the full peptide IDs in the workspace of your search engine step as idXML. You can view or export them to csv with (py)OpenMS.