Convert Proteomics Thermofischer Raw data to mzML open format

[proccaserra]

@ulo please see:
https://github.com/compomics/ThermoRawFileParser

the tool is described in the following manuscript (preprint)
https://www.biorxiv.org/content/10.1101/622852v1.full

it would help with the release of the dataset public and for future

[ulo]

Thanks for the input.
Yes, the currently community-accepted open format for proteomics is mzML.
In the past, I have used msconvert (http://proteowizard.sourceforge.net/tools/msconvert.html) for this conversion step. I think we should publish the open-format files in addition to the proprietary raw format.

[mcourtot]

@proccaserra sent info on tools - @sedlyarov will do the conversion and submit to proteomXchange. @ulo looking into accompanying metadata for submission

[ulo]

I submitted the proteomics data to the public ProteomeXchange repository, and made some interesting observations regarding requirements on file formats and metadata:

1. This repository requires not the raw files in an open format (which would be mzML; as also stated by @proccaserra) but the result files. As we used the ProteomeDiscoverer software, the proprietary result file format is *.pdResult. Fortunately, the software can also export the required mzID format.

2. Next to generic metadata on the whole data set (project title & description, keywords, sample & data processing protocol), they also require a number of more specific annotations on the sample and method:

   • species: NCBITAXON ontology
   • tissue: BTO / EFO ontology
   • instrument: MS ontology
   • cell type: CL / EFO ontology
   • disease: EFO / DOID ontology
   • quantification method: PRIDE ontology

3. Interestingly, the actual experimental factors are entered in a free text field, not requiring any structure or ontology.