Nesvilab / MSFragger

Ultrafast, comprehensive peptide identification for mass spectrometry–based proteomics
https://msfragger.nesvilab.org
100 stars 7 forks source link

Scan type HCD not recognized #235

Closed JB91451 closed 1 year ago

JB91451 commented 1 year ago

Dear all,

in order to use the very same peaklist files for different search engines, I had to convert some of my files to mzXML before searching them. When I use these files with MSFragger 3.5 I get a warning message during file checking and calibration saying: "Activation type HCD not recognized, activation filtering may be inconsistent! Please contact the developers to fix and/or disable activation type filtering" The files originate from a QExactive plus measurement in DDA mode and were converted via msconvert. Within the mzXML files, msLevel=2 scans have a tag 'activationMethod="HCD"' along with the precursor information. Also, the number of scans displayed during parameter optimization for each file seems valid. Therefore I would like to ask, whether it is save to ignore this warning (I have not mixed activation schemes etc.)?

Best regards, Juergen

Please find the used parameter file here. 'activation_types = HCD' had to be specified as without this information the searches immediately stopped.: PPutida_Proteogenomics_semiLysC_MSFragger.params.txt

Below is an example for a MS2 scan in the mzXML files:

<scan num="3702"
      scanType="Full"
      centroided="1"
      msLevel="2"
      peaksCount="366"
      polarity="+"
      retentionTime="PT1688.78S"
      collisionEnergy="27.5"
      lowMz="100.076606750488"
      highMz="1009.432556152344"
      basePeakMz="120.0816037"
      basePeakIntensity="5.7522012e05"
      totIonCurrent="2.891132e06"
      msInstrumentID="1">
  <precursorMz precursorScanNum="3700" precursorIntensity="8.67563701931e05" precursorCharge="3" activationMethod="HCD" windowWideness="3.0">334.5353245056</precursorMz>
  <peaks compressionType="none"
         compressedLen="0"
         precision="64"
         byteOrder="network"
         contentType="m/z-int">QFkE5yAAAABAjU1iwAAAA...</peaks>
</scan>
fcyu commented 1 year ago

Hi Juergen,

Can you send on of the mzXML file to us?

Thanks,

Fengchao

JB91451 commented 1 year ago

Hi Fengchao,

if you would open me a box folder, I can try to do so. However, the files are really large (8-12 Gb).

fcyu commented 1 year ago

Sure, here it is: https://www.dropbox.com/request/WG7nJrF13dPr22XpthE4

Best,

Fengchao

JB91451 commented 1 year ago

Thanks, it should upload. I also added the mzbin, search results (which seem good to me) and the fasta.

Best, Juergen

fcyu commented 1 year ago

Hi Juergen,

FYI, the dropbox folder is still empty. Are you still uploading the files?

Best,

Fengchao

JB91451 commented 1 year ago

Hi Fengchao,

thank you for informing me. There was a web-connection error with our server and thus the upload failed. Meanwhile it should be finished as I repeated it and got a confirmation.

Best, Juergen

fcyu commented 1 year ago

Hi Juergen,

Thanks for your files. I couldn't reproduce the warning message. Everything looks good from my test. Attached please find the log file.

Best,

Fengchao log.txt

JB91451 commented 1 year ago

Dear Fengchao,

Thank you for checking this. I still get the same error but as the number of spectra is the same as in your log file I guess I can simply ignore it?!

Best, Juergen

fcyu commented 1 year ago

I don't know if it is safe to ignore it. I think you need to troubleshoot from your side.

Best,

Fengchao

dpolasky commented 1 year ago

Hi Juergen, If you remove the activation_types=HCD parameter from the MSFragger parameters file (or change it to activation_types=all), the message should go away. Unless you need to filter in a file with multiple activation types, this parameter shouldn't be needed, and any unexpected behavior from trying filtering an mzXML file can be avoided. Best, Dan

JB91451 commented 1 year ago

Thank you both for your recommendations. I will try to run it with activation_types=all (I added the HCD as leaving this parameter blank did not work at the beginning) and try in parallel with the original raw files. However, I am quite optimistic that the results will be identical or close.

Thanks again, Juergen

fcyu commented 1 year ago

OK, I see the reason.

Dan, does it mean that we won't be able to support mzXML in the future? At least the activation_types parameter won't support it.

Best,

Fengchao

dpolasky commented 1 year ago

Yes for activation type filtering we probably won't support mzXML (unless there's a lot of demand for it), but can change it so that it doesn't try to filter rather than giving this scary error message

fcyu commented 1 year ago

Actually, the mzXML format has the activation method information:

<precursorMz precursorScanNum="839" precursorIntensity="3.544250343306e06" precursorCharge="2" activationMethod="HCD" windowWideness="3.0">324.172088534973</precursorMz>

I guess we can do something to recognize it.

Best,

Fengchao

dpolasky commented 1 year ago

yes the info is there in mzXML so we could support it, just need to test to find all the different ways it can be reported by different instruments/vendors

fcyu commented 1 year ago

I tested it with activation_types = HCD. MSFragger actually can recognize the activation type from mzXML format. And there are still no error messages. Then, I am not sure why @JB91451 got it,

Best,

Fengchao

JB91451 commented 1 year ago

Dear all,

I tried with activation_types = all over night and there is no warning anymore. As the files are identical to the ones produced with HCD filtering I think it was safe in my case to ignore the warning (at least in my case). Nonetheless I will use the files created without filtering to be sure.

Many thanks again, Best, Juergen