Discrepancy in results from running latest MaxQuant on Windows and Linux machine

[animesh]

Describe the bug
I tried running latest MaxQuant version 2.5.2.0 with the same parameters on a Linux and Windows machine but i see significant difference in results, for example the number of MSMS scatter (x-Windows, y-Linux)

This was not the case with previous version 2.4.14.0 where results look like

respectively.

I am wondering why this could be? Dotnet? and which results to report?

[bi-MPIB]

Thank you for posting this issue. We are now trying to repeat this bug. Could you please offer more details on what you are plotting here? e.g. What kind of data is it? From which output table, which column?

[animesh]

Thanks @bi-MPIB for looking into this. It is log2 of MSMS counts in proteinGroups.txt table for 6 timsTOF-pro data for a project i am working on currently. I am trying to move to the latest MaxQuant but noticed this so a bit unsure if i should? Since the data is sensitive so unfortunately i cannot share it... how does it look like for your data, i.e. running the same parameters on Linux and Windows and plotting log2MSMS values?

[bi-MPIB]

Actually, I can not reproduce the issue you have. Here the output of 3 Hela fractions runn in timsTOF from Windows and Linux are still amost the same in MQ2.5.2.0. Could you please double check your parameter files?

[animesh]

Thanks for sharing your experience @bi-MPIB , looks pretty similar to mine ? How does it look like with previous version, specifically 2.4.14?

PS: parameter file is same for Windows and Linux 👍