Thank you for a very interesting article, which I have found so useful it’s in my bookmarks.
Following your instructions I have converted a raw file to mzML then to mzXML and successfully started a MaxQuant search on it. The file had been had been acquired last August on an unlabelled sample.
However, a file run on the same instrument last month is causing MaxQuant to crash immediately after I hit start. There are a few minor differences between both files (the one working has a shorter run as it is a QC file; the one failing is a SILAC-heavy only file), but I cannot see how these could be causing issues. MaxQuant is able to search both of the original raw files.
In addition, I would like your opinion on a question I had: since it is possible to create pseudo-MSMS files from SWATH data using DIA Umpire, do you think it would be possible to then combine these with the MS1s from the original SWATH file to create a pseudo-DDA file containing both MS1 and MS2 that could be searched with MaxQuant? I understand this would beat the purpose of using a library, but could be useful for specific workflows.
Its hard to say what went the cuplrit could be, without some more details, e.g. did you use the -force_MQ_compatibility flag for FileConverter? Which MQ version did you use? (there might be changes in later versions?).
I updated the article with some new information regarding the mzXML library of MaxQuant. Maybe that helps.
In principle, the content of the Raw file should not matter, in terms of successfull conversion.
SWATH: The workflow should be feasible. Any particular reason you’d prefer MaxQuant/Andromeda over other search engines?
Combining results (iProphet & Co) might be very interesting. Merging pseudo MS2 with MS1 is a 2-node workflow with OpenMS FileFilter and FileMerger, so very easy.
I tried to convert data from Bruker (ion-trap) to be analyzed by MaxQuant 1.6.7, but could not find the file converter on your link above. I downloaded the OpenMS full version from the link below.
Then I used command-line version of file converter in bin folder but the result filed still crashed MaxQuant at the “test raw file” step. (I tried all combinations mzML, mzXML, mzData but still did not work.
Is the converter I am using different from what you were using in the article? Can you provide a download to the converter in your article?
Here are a few other changes to the mzxml that may be important for formatting mzXML that is acceptable to MaxQuant. I haven’t had time to verify them, but writing them down in case some other poor soul needs ideas on how to fix cryptic assertions from MaxQuant:
– normalize the intensities of each spectrum to, say, 1000 max. This avoids overflow.
– scan numbers are limited to a 32 bit int and should be consecutive
Thorough tutorial on the topic!
Thanks!
Thank you for a very interesting article, which I have found so useful it’s in my bookmarks.
Following your instructions I have converted a raw file to mzML then to mzXML and successfully started a MaxQuant search on it. The file had been had been acquired last August on an unlabelled sample.
However, a file run on the same instrument last month is causing MaxQuant to crash immediately after I hit start. There are a few minor differences between both files (the one working has a shorter run as it is a QC file; the one failing is a SILAC-heavy only file), but I cannot see how these could be causing issues. MaxQuant is able to search both of the original raw files.
In addition, I would like your opinion on a question I had: since it is possible to create pseudo-MSMS files from SWATH data using DIA Umpire, do you think it would be possible to then combine these with the MS1s from the original SWATH file to create a pseudo-DDA file containing both MS1 and MS2 that could be searched with MaxQuant? I understand this would beat the purpose of using a library, but could be useful for specific workflows.
Its hard to say what went the cuplrit could be, without some more details, e.g. did you use the -force_MQ_compatibility flag for FileConverter? Which MQ version did you use? (there might be changes in later versions?).
I updated the article with some new information regarding the mzXML library of MaxQuant. Maybe that helps.
In principle, the content of the Raw file should not matter, in terms of successfull conversion.
SWATH: The workflow should be feasible. Any particular reason you’d prefer MaxQuant/Andromeda over other search engines?
Combining results (iProphet & Co) might be very interesting. Merging pseudo MS2 with MS1 is a 2-node workflow with OpenMS FileFilter and FileMerger, so very easy.
cheers
Chris
Dear Chris,
Thank you for your nice article.
I tried to convert data from Bruker (ion-trap) to be analyzed by MaxQuant 1.6.7, but could not find the file converter on your link above. I downloaded the OpenMS full version from the link below.
https://www.openms.de/download/openms-binaries/
Then I used command-line version of file converter in bin folder but the result filed still crashed MaxQuant at the “test raw file” step. (I tried all combinations mzML, mzXML, mzData but still did not work.
Is the converter I am using different from what you were using in the article? Can you provide a download to the converter in your article?
Hi Jaran,
sorry for replying so late. The comment got stuck in review.
Can you point me to the data you were trying to convert? I can have a look.
cheers
Chris
Here are a few other changes to the mzxml that may be important for formatting mzXML that is acceptable to MaxQuant. I haven’t had time to verify them, but writing them down in case some other poor soul needs ideas on how to fix cryptic assertions from MaxQuant:
– normalize the intensities of each spectrum to, say, 1000 max. This avoids overflow.
– scan numbers are limited to a 32 bit int and should be consecutive