How to identify/quantify proteins with unique peptides only

[PatrickT2303]

Hi all,

I have a fairly simple question, but I am not sure if I found the right settings and would appreciate any help.

I am running DIA data (from Astral 40 min gradients) of different genotypes in Physcomitrium (total protein samples). The mutations are in two genes that share multiple peptides with each other. Therefore, I would like to set up a DIA-NN run with my samples, but for the identification/quantification I only want unique peptides to be used.

I am using DIA-NN V2.1.0 GUI version on a windows computer, with library free search (uniprot fasta file). I checked the how-to section and found that I would have to change the Scoring from "Generic" to "Proteoforms"- is that correct? Are there any other changes that I have to make besides this?

Thanks for any help!
Best,
Patrick

[vdemichev]

Hi Patrick,

Yes, Generic to Proteoforms. If genes are named differently for paralogues, please just use Genes.MaxLFQ.Unique, otherwise (same gene names) please also add --ids-to-names.

Best,
Vadim

[PatrickT2303]

Hi Vadim,

Thank you very much for your answer. I was able to do the run in proteoforms scoring mode, but I am not entirely sure what you mean with "just use Genes.MaxLFQ.Unique...". The fasta file I use has two entries, one for each gene/protein, so they are named differently. In case I miss something, can you please specify what you mean?

DIA-NN also finished the "proteoform" run, and when going over the report.pr.matrix file I could find for one of the two proteins only entries/peptides that are not shared with other proteins (same was true for the Generic run), so I guess unique. However, for the other protein I get, besides unique entries, also entries/peptides that were assigned to one protein group, but contain two protein IDs (from those two proteins that are highly similar). So my question is right now, are those shared peptides also accounted for in the overall protein abundance/intensity calculations for that particular protein (in report.pg.matrix) in proteoform scoring mode, or are they NOT included, but DIA-NN is still listing those identified shared peptides in the report.pr.matrix file?

Thanks,
Patrick

[vdemichev]

just use Genes.MaxLFQ.Unique

I mean the respective column in the main .parquet report.

The fasta file I use has two entries, one for each gene/protein, so they are named differently

So if genes are named differently, then Genes.MaxLFQ.Unique will be correct :)

The pg_matrix will also used shared peptides, i.e. not suitable in this case.

[PatrickT2303]

Hi Vadim,

Thanks for the information! I was able to look up the data from the main.parquet file, but the overall data output is unfortunately not compatible with my downstream processing ( I am using Perseus). It looks to me as if the LFQ values are for a specific peptide for a given protein rather than an "overall" LFQ value for that particular protein. I was hoping the output would be like the one in the pg_matrix file. Is there an (easy?) workaround to get this kind of information (meaning "unique" LFQ values for a protein group for each sample)?

Thanks for all your help!

[vdemichev]

Hi Patrick,

obtain the protein group specific LFQ values

PG.MaxLFQ is what you have in mind?

Best,
Vadim

[PatrickT2303]

Yes, that is exactly what I am hoping for!

[vdemichev]

So all clear?

[PatrickT2303]

Not entirely, you mentioned earlier in this thread that "The pg_matrix will also used shared peptides, i.e. not suitable in this case.". I am only looking for a pg matrix output with LFQ intensities for all protein groups that have been identified with UNIQUE peptides only.

[vdemichev]

i.e. not suitable in this case

Genes.MaxLFQ.Unique in combinations with --ids-to-names and Proteotypicity set to Isoforms.