Take Unimod modification compositions from psims.controlled_vocabulary.unimod

[vrkosk]

Currently, DeepLC hardcodes Unimod modification compositions in unimod_to_formula.csv. As new modifications are added, this file will grow out of date. It would be better to take modification compositions from a central place like psims.controlled_vocabulary.unimod.

[RobbinBouwmeester]

Actually the documentation is out of data here. It currently uses:

https://github.com/compomics/psm_utils

Where most of the parsing (as far as I know) of the modifications is done with pyteomics.

When adding new modifications you can use the ProForma composition encoding:

SEQUEN[Formula:C12H20O2]CE

[RobbinBouwmeester]

Will update the documentation soon.