search

smith-chem-wisc / MetaMorpheus

Proteomics search software with integrated calibration, PTM discovery, bottom-up, top-down and LFQ capabilities
MIT License
90 stars 45 forks source link

Metamorpheus crashes during Quantification (not responding) #1492

Closed Peer2011 closed 5 years ago

Peer2011 commented 5 years ago

Dear rmillikin, I would like to analyze 38 Files of Q-Exactive HF. Calibration works perfectly. However, when I search files and would like to do LFQ with normalization and match between runs, after classic search RAM usage ramps up to 12,5 GB during the quantification process and the program stops responding. However, it still calculates somehow but I can't do anything anymore. Its running for two days in quantification now. I used the two variable modifications for deamidation besides standard settings and run metamorpheus with the most updated Thermo raw file reader. mzml files are generated during calibration process. So this looks all fine. Do I have too little RAM (I thought 16 GB should be enough)? Any help would be very much appreciated because I think that your software could be become a viable alternative to MaxQuant which we routinely use currently. If you need further information. Please let me know.

Best regards, Peer

rmillikin commented 5 years ago

Hi Peer,

Matching between runs is a memory hog right now, unfortunately. Fixing this has been on my to-do list for a while. I'm hoping to have time in the next week or two to take a look at it. It'll hopefully be in the next release - I'll try to keep you posted. Thanks for the crash report!

rmillikin commented 5 years ago

This has proven to be more difficult than I thought, but I've got it mostly working in my test code. I'm wrapping it up now and it will be in the next release. Stay tuned :) apologies for the delay.

Peer2011 commented 5 years ago

Really looking forward!!!

Von meinem iPhone gesendet

Am 10.11.2018 um 13:05 schrieb Robert Millikin notifications@github.com:

This has proven to be more difficult than I thought, but I've got it mostly working in my test code. I'm wrapping it up now and it will be in the next release. Stay tuned :) apologies for the delay.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

rmillikin commented 5 years ago

This is finally out in a new release. If you still encounter RAM issues, it's probably due to MetaMorpheus search and not FlashLFQ quantification. We're still planning on making RAM improvements to the search in the future. Can you give the new version a try?

Thanks, Rob

Peer2011 commented 5 years ago

Of course.

Best, Peer

Von meinem iPhone gesendet

Am 26.11.2018 um 19:54 schrieb Robert Millikin notifications@github.com:

This is finally out in a new release. If you still encounter RAM issues, it's probably due to MetaMorpheus search and not FlashLFQ quantification. We're still planning on making RAM improvements to the search in the future. Can you give the new version a try?

Thanks, Rob

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.