Open searches constantly out of memory about msfragger HOT 10 CLOSED

How big is humanUniref100.fasta database? How many sequence entries are in there? 24G is not a lot of memory since you want to do phospho+ variable on K You can try to use a smaller database like UniProt, reviewed sequences only. You can restrict peptide length to 8-40 amino acids (instead of 7-50). This will help too. MSFragger takes advantage of indexing to get its amazing speed, so if you do not have enough memory to build the index you may not be able to use MSFragger. You can also contact me directly (not via github) and send your mzML file and the sequence database, and I can test on my computer that has 64G to see if it is really the memory issue. Best, Alexey From: wzhao6898 [mailto:[email protected]] Sent: Tuesday, September 18, 2018 12:57 PM To: chhh/FragPipe <[email protected]> Cc: Nesvizhskii, Alexey <[email protected]>; Comment <[email protected]> Subject: Re: [chhh/FragPipe] Open searches constantly out of memory (#75) I have 24G and 8 cores on my windows 7 64 bit PC, and actually -17 QC and -18E variable modifications are commented out, I only searching: variable_mod_03 = 79.96633 STY variable_mod_04 = 196.13683 K Thanks — You are receiving this because you commented. Reply to this email directly, view it on GitHub<https://github.com/chhh/FragPipe/issues/75#issuecomment-422469740>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AhDGe2hBEbXe3NJ3XlNF9pTViorrhIF5ks5ucSX4gaJpZM4WuQIW>. ********************************************************** Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues

Are you running FragPipe? There are PeptideProphet, ProteinProphet, and FDR filter command you should also run. The resulting PSM and protein tables are then filtered to 1% protein FDR and 1% PSM FDR. Are you doing non-standard analysis that requires hyperscore (or expectation value) filtering instead if FDR-based filtering?

I suggest you use expect value (instead of hyper score), and something like e-value < 0.01 But ii is really dataset dependent - -the very reason the field moved away from using arbitrary score thresholds for any search engine, MSFragger would not be any different. Best, Alexey From: wzhao6898 [mailto:[email protected]] Sent: Wednesday, September 19, 2018 12:53 PM To: chhh/FragPipe <[email protected]> Cc: Nesvizhskii, Alexey <[email protected]>; Comment <[email protected]> Subject: Re: [chhh/FragPipe] Open searches constantly out of memory (#75) I'm actually running MSFragger GUI 6.0, and I did run PeptideProphet and proteinprophet, I was just looking to see if there is an empirical hyperscore threshold observed. Thanks, David — You are receiving this because you commented. Reply to this email directly, view it on GitHub<https://github.com/chhh/FragPipe/issues/75#issuecomment-422877210>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AhDGe0IIoeXtKXSgbDpRAINa0fnT0BW-ks5ucnZjgaJpZM4WuQIW>. ********************************************************** Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues

We have a new version that solves the ‘out of memory’ problem. If you are an academic/non-profit scientists (so you can get free version), please contact me and we will give you a new version to test. Best Alexey From: wzhao6898 [mailto:[email protected]] Sent: Wednesday, September 19, 2018 12:53 PM To: chhh/FragPipe <[email protected]> Cc: Nesvizhskii, Alexey <[email protected]>; Comment <[email protected]> Subject: Re: [chhh/FragPipe] Open searches constantly out of memory (#75) I'm actually running MSFragger GUI 6.0, and I did run PeptideProphet and proteinprophet, I was just looking to see if there is an empirical hyperscore threshold observed. Thanks, David — You are receiving this because you commented. Reply to this email directly, view it on GitHub<https://github.com/chhh/FragPipe/issues/75#issuecomment-422877210>, or mute the thread<https://github.com/notifications/unsubscribe-auth/AhDGe0IIoeXtKXSgbDpRAINa0fnT0BW-ks5ucnZjgaJpZM4WuQIW>. ********************************************************** Electronic Mail is not secure, may not be read every day, and should not be used for urgent or sensitive issues