Add ProForma 2.1 Peptidoforms to FragPipe output

[mobiusklein]

We are working on wrapping up the ProForma 2.1 peptidoform notation update (https://www.psidev.info/proforma). It would be great if we could add a column to the psm.tsv, ions.tsv, and/or peptide.tsv output tables with the ProForma notation for the identified peptides so that there is a standardized way to extract the peptidoforms. Right now, this information is spread across at least three columns (Peptide, Assigned Modifications, Observed Modifications, various localization columns). There are several converters out there for translating some version of FragPipe's outputs to ProForma 2 in the wild, but each supports only a subset of features needed for their use-cases.

I think @douweschulte mentioned that the chemical formula definition in glycans was actually a feature added to the 2.1 update based upon a discussion with someone associated with FragPipe.

What, if anything, can we do to help facilitate supporting this? What code, references, or other material if any, could be contributed to help make this work?

[mobiusklein]

Per discussion today:

I'll start off small and provide either a Rust or Java tool to implement a ProForma "building" API that can be fed from the four or so FragPipe columns from psms.tsv or precursors.tsv, and the relevant information will be rolled up to the later tables by a downstream step. The basics of:

1. Unimod PTMs
2. Delta masses
3. Glycan compositions, including custom monosaccharides
4. GNOme accessions (1:1 map to GlyTouCan accessions 99.9% of the time or better)
5. Localization confidence (if available in some test data)

Fixed modifications will still be written inline without using ProForma's fixed modification rules, nor will I try to guess at stable isotope labeling yet.

[dpolasky]

Sounds good, thanks Joshua! Let us know if you need any test data or etc.

Best,
Dan