PFAM domains for all protein Isoforms from same protein

Question

[Michi]

Hi

Is it possible to get PFAM domain mapping for different isoforms of the same protein?

At Uniprot for example they report all positional data relative to their canonical isoform.

At PFAM, I found this proteome-file: ftp://ftp.ebi.ac.uk/pub/databases/Pfam/releases/Pfam27.0/proteomes/9606.tsv.gz. This file nevertheless seems to link the Uniprot identifiers - without specifying any isoform

So I am a bit stuck and don't know how to proceed. Or maybe I misunderstood something(s)?

Thanks for any additional info/help

Answer 1

[cdsouthan]

Not from UniProt because I don't think the isoforms are run through Pfam. Running each of them them yourself would be hard work and I'm not sure how useful that would be. Isoforms with a missing exon (i.e. most of them) would just gap the global allignments that are compiled from the canonical (all exon) forms

[Michi]

So what would you do if you'd like to map the domains from the canonical transcript to the isoforms?

[cdsouthan]

If you could expand on the biological question you are trying to answer I might be able to help more. On a a coordinates basis sure - you can map domain markup across to residue positions in isoforms but this is not biologically valid. If you delete a sizable exon chunk out of a Pfam domain (or any other InterPro signature in fact) that part of the protein is probably not going to fold properly - regardless of domain boundaries.