search

compomics / peptide-shaker-2.0-issue-tracker

Issue tracker for the beta release of PeptideShaker 2.0
Apache License 2.0
0 stars 0 forks source link

Not enough representative modification sites found #44

Open MarcIsak opened 4 years ago

MarcIsak commented 4 years ago

Hi Harald,

I am using PeptideShaker 2.0-beta for Windows to process search results from Comet and X!Tandem in SearchGUI 4.0.0-beta. I have searched 7 RAW-files with sizes around 1 GB.

Unfortunately, there is a step in which processing in PeptideShaker always stops. It is at the Scoring Peptide Modification Localization. It is when 85% of the processing has been complete, see attached image. error_peptshaker

My thought is that it could be related to memory issues.

I am using 6 cores with 24 GB allocated for Java 12.0.1 (JDK, 64-bit). So it should be 4 GB RAM available for each core. Still, I wonder if it is enough? How much memory do you allocate for each core when you process btw?

I just thought it could be interesting to mention, since it also could be a bug.

Best,

Marc

hbarsnes commented 4 years ago

Hi Marc,

There is probably an undetected bug in the PTM scoring. Can you share the PeptideShaker log file? You'll find it via the Welcome dialog > Settings & Help > Help > Bug Report.

Best regards, Harald

MarcIsak commented 4 years ago

Hi Harald,

I tried to process new data in PeptideShaker 2.0.0-beta, and it stopped at Scoring Peptide Modification Localization again:

pept_shaker_v200_error

This time, I can provide you with the error log file, see attached.

PeptideShaker 2.0.0-beta log.txt

hbarsnes commented 4 years ago

Hi Marc,

Ok, so there is a bug in the localization of the PTMs. Which search engines are you using? And which variable PTMs?

And could you try running without PTM scoring? I.e. set the "Probabilistic Score" to "None" and "Confident Sites" under "Site Alignment" to "No" in the "PTM Localization" settings dialog.

Best regards, Harald

MarcIsak commented 4 years ago

Hi Harald,

I used Comet, X!Tandem and MS Amanda. The searches went fine, even with MS Amanda this time.

I will try to run another analysis later today, where I have disabled the things mentioned above. I will get back to you after this.

Best,

Marc

MarcIsak commented 4 years ago

...and I only used Oxidation on Methionine as variable mod.

Best,

Marc

MarcIsak commented 4 years ago

Hi Harald, I managed to import the SearchGUI output.

Still, I cannot save the project locally as a psdb-file and load it (but I opened a new issue for this).

hbarsnes commented 4 years ago

Hi Marc,

I used Comet, X!Tandem and MS Amanda. The searches went fine, even with MS Amanda this time.

Could you also try loading just the X! Tandem results?

Best regards, Harald

hbarsnes commented 4 years ago

Hi Marc,

I'm having a hard time with the memory/database issues. Made some progress, but seem like I need help from the two other developers that were in charge of implementing the new database backend.

However, if only loading the X! Tandem data I am able to both load, save and reopen. Perhaps you can use that as a temporary fix until we're able to solve the other issues?

Best regards, Harald