search

compomics / peptide-shaker

Interpretation of proteomics identification results
http://compomics.github.io/projects/peptide-shaker.html
48 stars 19 forks source link

PeptideShaker importing error from SearchGUI output #512

Open fengx90 opened 1 year ago

fengx90 commented 1 year ago

Dear Compomics team,

Thanks very much for a such great tool for proteomics study. Much appreciated! I have analyzed several files in SearchGUI using multiple search engines and intended to validate/integrate the output in PeptideShaker, but unfortunately, Peptideshaker failed the automatic processing. I have tried importing the search output manually into Peptideshaker, but it crashed more or less at the same step "Selecting leading proteins, inferring peptide and protein inference status". Then the importing was canceled.

I hope you can help me out with this. Thanks very much in advance!

Feng

hbarsnes commented 1 year ago

Hi Feng,

Can you please share the PeptideShaker log file? You can find it via the PeptideShaker Welcome dialog > Settings & Help > Help > Bug Report.

You may also look into increasing the amount of memory given to PeptideShaker. This is done via the PeptideShaker Welcome dialog > Settings & Help > Settings > Java Settings > Memory.

Best regards, Harald

fengx90 commented 1 year ago

Hello Harald,

Thanks for the quick reply. Yes, I exported the bug report and attached here. PeptideShaker 2.2.8 log.zip

Yes, I also tried to increase the memory for Java, as I read your reply to a previous issue. At the moment I am using 110 GB for Java (out of 120GB) and 20 (out of 24) or 40 (out of 48) cores depending on the PCs.

Looking forward to your reply!

Best, Feng

hbarsnes commented 1 year ago

Hi Feng,

I'm afraid you are experiencing several issues with the backend for which there is no known fix at the moment. We are however working on refactoring this part of the code which should hopefully solve the issues. I cannot say how long it will take to complete this work though, as it is more complicated than we had hoped.

In the meantime you can perhaps try to reduce the number of search engines, as this could result in the issues not happening (or at least not happening as frequently). Which search engines are you using at the moment? Generally, they all provide more or less the same results, hence you can get by with using just one or two of them without it negatively effecting the results. The added bonus is that this will also greatly increase the speed of the processing.

The only other thing that you may try is to run PeptideShaker using a single thread/core. This is done via the Project Settings at the bottom of the New Project dialog. Note however that this will make the processing much slower.

I'm sorry that we cannot provide a better fix at the moment.

Best regards, Harald