[autoiRT] First pass sample from commonly seen / known endogenous eukaryotic peptides

[singjc]

When sampling for iRTs from the input PQPs, especially for in-silico libraries, could we add a priority inclusion list of peptides that we try finding first to sample for, then sample the rest once we don't find then anymore.

There are a set of conserved eukaryotic peptides that have been compiled into a list of CiRTs. Could we use this as a first pass inclusion list in the PQP iRT sampling stage. This would hopefully increase the likelihood of the sampled iRTs from the in-silico library at least being potentially detected when performing the calibration.

Should we ship the cirtkit.txt/TraML in share/OpenMS/CHEMISTRY. @timosachsenberg, do you think we can/should include these iRT files (biogynosis spiked-in iRT and CiRT) with OpenMS?

OpenMS PR #8373

Ref: https://www.mcponline.org/article/S1535-9476(20)32633-5/fulltext

[timosachsenberg]

Not an expert, but I think it would be useful.
Adding to chemistry would probably be the most fitting location.
We could even revisit this list if it needs to be extended (e.g., from existing quantms results)

[singjc]

Not an expert, but I think it would be useful. Adding to chemistry would probably be the most fitting location. We could even revisit this list if it needs to be extended (e.g., from existing quantms results)

I was actually thinking about that as well! It would be great if we could update the CiRT list with existing quantms results. I was thinking if we could also use the existing quantms results to generate a new updated Pan Human library, to see how that compares against a large in silico library.

[timosachsenberg]

@ypriverol do you have easy access to DIA-NN frequently identified peptides/transitions?

[ypriverol]

you mean predicted library from Human Uniprot?

[timosachsenberg]

and statistics which peptides have been identified / quantified by DIA-NN over many samples. Basically, what is always there (e.g. housekeeping proteins)

[ypriverol]

for an specific dataset, yes.

[timosachsenberg]

would it be possible (e.g. for a student) to create aggregated statistics over many datasets to see which are the e.g. top 100 peptides always observed in human data?