Hi everyone, I've been searching for a way to retrieve protein isoforms sequences programmatically but unfortunately didn't succeed so far. After parsing the whole Uniprot Trembl and Swissprot xml files I've seen that the sequences are not included, just the isoforms definition plus some extra information. Do you know if there's any kind of web service for it? Thanks in advance,
Pablo Pareja
http://www.uniprot.org/uniprot/<ANY_UNIPROT_ISOFORM_ID_THAT_YOU_HAVE>.fasta
replace the part in CAPITALS with isoform accesions that you want to find.
e.g.
http://www.uniprot.org/uniprot/P05067-2.fasta
This won't work for the cases where the isoform id is the canonical sequence and there is just one sequence in the record.
Should be okay week after next for these cases as well.
Hi jerven. Welcome to Biostars! Your first link is broken and needs some mending. Cheers!
Hi jerven. Thanks for your answer, I assume then that this url-pattern should work from mid-February on. I have to ask anyways, isn't there a way for retrieving more than just one isoform sequence at a time? e.g. generating a multi-fasta file with several isoform sequences included? Cheers
Check out this section of the FAQ. Note that although it mentions they can be downloaded here , the link actually goes to the documentation of Paul Kersy's varsplic.pl program which is usually found here
I didn't realized so far that there's a file including every isoform sequence in Uniprot downloads site:
Maybe this is not the best option for some people willing to use a web service instead but in my case it fits my requirements perfectly. (Just with a really simple Java program I can extract all the information I need from any protein isoform.)
Heads up: in 2018 this file contains only non-canonical splice variants (not all isoforms). Canonical isoforms are stored in ftp://ftp.uniprot.org/pub/databases/uniprot/current_release/knowledgebase/complete/uniprot_sprot.fasta.gz which is an order of magnitude bigger.
Log in to answer this question.
