elixir-proteomics-community / sdrf_convert

3 stars 0 forks source link

Access modifications #1

Closed julianu closed 1 year ago

julianu commented 1 year ago

The modifications are in SDRF per default given as a simple name, like "Oxidation", and the site, like M.

While the site is usually ok for most search engines, the simple name in many engines is not, but a mass shift like "+15.9949" is needed ( e.g. for XTandem, there is no naming of modificatiosn at all, only mass shifts).

I saw @di-hardt already has some annotations in is code to use pyteomics to map this. Before making dual efforts: is there already any progress, or should I go ahead and make an implementation into the abstract_converter? I think many parsers will need any kind of this functionality in the end.

veitveit commented 1 year ago

Hi @julianu I compiled information about these a while ago: https://github.com/wombat-p/WOMBAT-Pipelines/blob/dev/assets/unimod2searchgui_mapping.tsv Maybe that helps.

di-hardt commented 1 year ago

The SDRF should include the modification accession I thinks. pyteomics provides an easy to use interface for accessing the unimod using the accession: https://pyteomics.readthedocs.io/en/latest/api/unimod.html For custom modification I thought about an YAML / JSON file for lookups. I need to implement it for the Comet converter as well.

julianu commented 1 year ago

I created a PR #2, which adds pyteomics and a simple query into the abstract class.

julianu commented 1 year ago

As the #2 did not work, it should be in #3

di-hardt commented 1 year ago

Merged