createMSP error

[bernt-matthias]

We have the following error for createMSP:

Error in msPurity::createMSP(pa, msp_file_pth = file.path(opt$out_dir,  : 
  object 'pa' not found

input comes from xcms findChromPeaks (xcmsSet) Chromatographic peak detection (Galaxy Version 3.4.4.0)

I guess its an incompatibility with this newer

Was also wondering if the input data type of the tool should be: rdata.xcms.findchrompeaks (which seems to be xcms.set output type in 3.x) or rdata.xcms.raw (which seems to be xcms.set output type in 2.x)

[Tomnl]

hi @bernt-matthias ,

The createMSP tool should use the purityA object as input (after frag4feature and any other processing).

e.g.

• Purity assessments
  • (mzML files) -> purityA -> (pa)
• XCMS processing (old functions used here but should work with xcms>=3
  • (mzML files) -> xcms.xcmsSet -> xcms.merge -> xcms.group -> xcms.retcor -> xcms.group
    -> (xset)
• Fragmentation processing
  • (xset, pa) -> frag4feature -> filterFragSpectra -> averageFragSpectra -> createMSP -> (MSP file)

The documentation in the Galaxy tool is a bit lacking but the R documentation is a bit more in depth. See https://bioconductor.org/packages/release/bioc/manuals/msPurity/man/msPurity.pdf

[bernt-matthias]

Hi @Tomnl. Thanks for the response. At the moment we only want to use msPurity functionality for creating msp files from mzML files. What would be the preferred workflow for this?

However, I would suggest to use more specific Galaxy data types than rdata to help users to select inputs of the correct type.

[Tomnl]

What MS/MS do you want? i.e.

• Every single MS/MS scan in the mzML file?
• Every single MS/MS scan associated with an XCMS grouped feature?
• Averaged MS/MS per XCMS grouped feature?

I will make an issue for the data types.

PS - I am making some updates at the moment that might fix some bugs. Are you using the testtoolshed or working directly from the github code?

[chufz]

Hi Tomni,
for the moment we would prefer to get Every single MS/MS scan associated with an XCMS grouped feature, extract the msp, show how strong they variate among the sample set and use them further for use in metfrag / cfm

[Tomnl]

Hi @chufz

The minimum workflow with msPurity for that would be the following:

• Purity assessments
  • (mzML files) -> purityA -> (pa)
• XCMS processing (old functions used here but should work with xcms>=3
  • (mzML files) -> xcms.xcmsSet -> xcms.merge -> xcms.group -> (xset)
• Fragmentation processing
  • (xset, pa) -> frag4feature -> createMSP -> (MSP file with all the fragmentation)

For createMSP - make sure the option for "How to choose fragmentation spectra" is set to "all". This will create an MSP file with all the MS/MS associated with each XCMS feature.

However, from my experience it is better to at least average the spectra (and in some cases perform some basic filtering)

(xset, pa) -> frag4feature -> filterFragSpectra -> averageFragSpectra -> createMSP

[Tomnl]

Could I check as well, are you installing directly from the testoolshed or from the github repo?

[bernt-matthias]

I installed for testing directly from the repo, I'm at the latest commit 4059acd

[Tomnl]

Thanks @bernt-matthias, I am planning to merge to some updates from #15 by the end of the day. I will let you know when they are merged with stable and passing.

[bernt-matthias]

Thanks

[Tomnl]

#15 has been merged now

[bernt-matthias]

updated. thanks