How to distinguish protein isoforms using biomaRt?

Question

[richierocks]

I want to retrieve the canonical sequence for some proteins from ensembl, based upon their UniProt IDs.

library(biomaRt)

ensemblMart <- useMart(
  biomart = "ensembl",
  dataset = "hsapiens_gene_ensembl"
)

(result <- getBM(
  attributes = c("uniprot_swissprot", "peptide"),
  filters    = "uniprot_swissprot",
  values     = "P31749",
  mart       = ensemblMart,
  bmHeader   = TRUE
))

This particular protein has two isoforms, both of which are returned. I only care about the canonical form, so how can I distinguish the isotopes?

I suspect that I need to return another of the attributes from the mart

listAttributes(ensemblMart)

but I don't know which one.

Answer 1

[Emily]

There is no canonical defined in BioMart.

[richierocks]

Thanks. Just to be clear: are there any other ways of distinguishing isoforms?

[Emily]

You could perhaps use the APPRIS principal isoform, which is available from BioMart.

Answer 2

[shelly.deforte]

If you are trying to retrieve protein sequences using UniProt IDs, why not just fetch them directly from UniProt using something like urllib2 on Python (I'm sure there's a way to do this in R as well)? Or even just wget on linux, i.e., wget -nv http://www.uniprot.org/uniprot/O53166.fasta. UniProt will return the canonical sequence.