I like the idea of ProteomeXchange, which is a public archive for sharing mass spectrometry-based proteomics data. However, the current tool for preparing a submission of Trans-Proteomic Pipeline-processed material, "idconvert", does not seem to be functional (see discussion thread linked below). I currently perform searches using the version of X! Tandem that is bundled with the Trans-Proteomic Pipeline. Am I missing something? Are there other ways to construct conforming files for upload?
https://groups.google.com/forum/#!topic/spctools-discuss/NqJEjWJju4Y
I decided to wait a few weeks to see if updates to the "idconvert" program solved my problem, but no luck so far for my particular pipeline (the latest version I've tried is the one included with ProteoWizard 3.0.6867). Furthermore, going the mzIdentML route requires peak lists, which are not saved when using X! Tandem through the Trans-Proteomic Pipeline (TPP) with default parameters. Currently, I think the best option is to convert X! Tandem/TPP output files to PRIDE xml using the PRIDE converter. This requires spectra to be included in the X! Tandem/TPP output file (by adding "yes" to the parameters file to override the default), thus dealing with the missing peak list problem. The final submission to ProteomeXchange would include PRIDE xml results files made from the X! Tandem/TPP output, original raw data files, and PeptideProphet files provided as "other" files. Classification of the PeptideProphet files as "other" probably makes them inaccessible to big data projects, but they would be there in case anyone wanted to recreate the analysis.
For the record, the above approach worked and the data set is now available from the EBI PRIDE repository.
http://www.ebi.ac.uk/pride/archive/projects/PXD001528
- Badges
- Science method