Open hgbrian opened 1 day ago
Hi @hgbrian:
First of all, thanks for using quantms, I will try to do my best to help you here.
I will try to look into the details of the segmentation fault for the Comet. I will keep you updated. I must say I run Comet in docker and singularity with no issue in Linux. But I will try to reproduce the error.
The DDA pipeline by default, does not add decoys to the database, then you should add them manually or tell the pipeline to add it by using the parameter --add_decoys
.
I don't think the DDA pipeline (comet, etc) will work with DIA data. However, I do think this is a DDA dataset by reading the manuscript.
Description of the bug
First, to get the pipeline to not segfault on comet.exe, I had to update
'biocontainers/openms-thirdparty:3.1.0--h9ee0642_1'
tobiocontainers/openms-thirdparty:3.2.0--h9ee0642_4
. There is an old issue https://github.com/bigbio/quantms/issues/270 that references this so maybe the problem came back?I am trying to get a basic DIA pipeline to run and find some peptides. I have a DIA raw file (and its mzML) that I want to search for peptides in. I have tried the raw, the mzML, and various fast files. They all fail in the same way.
Since this was failing I tried one of the example files from
proteomics-sample-metadata
,PXD000396
and the 20k uniprot human protein fasta. This also fails in the same way as my original file, so I think this is a bug? (Or I am misunderstanding something very basic!)The basic problem seems to be that percolator cannot find decoys, which I think stems from the fact that comet does not find any peptides (so the idXML is empty of peptides.)
Should comet even be running for a DIA pipeline? (I have also tried setting -
-labelling_type "label free sample" --acquisition_method dia
to force DIA)Command used and terminal output
Relevant files
PXD000396.sdrf.tsv.zip human_proteome.fasta.zip
System information