Isoforms Missing In Uniprot Isoforms Fasta File

Question

[Pablo Pareja]

Hi,

I just ran into a problem regarding isoforms specified in Uniprot XML files which are not present in the 'every' isoform fasta file available in Uniprot.

One example would be the protein: Q8IYH5 where we can find the following isoforms:

<comment type="alternative products">
<event type="alternative splicing"/>
<isoform>
<id>Q8IYH5-1</id>
<name>1</name>
<sequence type="displayed"/>
</isoform>
<isoform>
<id>Q8IYH5-2</id>
<name>2</name>
<sequence type="described" ref="VSP_025511"/>
</isoform>
<isoform>
<id>Q8IYH5-3</id>
<name>3</name>
<sequence type="described" ref="VSP_025509 VSP_025510"/>
<note>No experimental confirmation available.</note>
</isoform>
<isoform>
<id>Q8IYH5-4</id>
<name>4</name>
<sequence type="described" ref="VSP_025512 VSP_025513"/>
<note>No experimental confirmation available.</note>
</isoform>
</comment>

Most of them can be found in the general isoform fasta file, however the isoform with ID: Q8IYH5-1 is not present in the file.

Do you know if this is a bug/isolated case? Is there any reason/relationship between the ones missing here?

Thanks in advance,

Pablo

[Elisabeth Gasteiger]

Hi Pablo, the isoform fasta file (uniprotsprotvarsplic.fasta) only contains additional isoforms (see ftp://ftp.uniprot.org/pub/databases/uniprot/current_release/knowledgebase/complete/README.varsplic) and not the "canonical" sequences, i.e. those marked <sequence type="displayed"/>

The isoform file needs to be combined with the file uniprotsprot.dat (or uniprotsprot.fasta) to obtain all isoform sequences.

See also What is the canonical sequence? http://www.uniprot.org/faq/30 How to retrieve sets of protein sequences? http://www.uniprot.org/faq/38

Elisabeth

Answer 1

[Elisabeth Gasteiger]

Sorry, this is my first post - I just posted my reply as a comment by mistake.

Here you go again:

The isoform fasta file (uniprot_sprot_varsplic.fasta) only contains additional isoforms (see ftp://ftp.uniprot.org/pub/databases/uniprot/current_release/knowledgebase/complete/README.varsplic) and not the "canonical" sequences, i.e. those marked

<sequence type="displayed"/>

in the xml.

The isoform file needs to be combined with the file uniprot_sprot.dat (or uniprot_sprot.fasta) to obtain all isoform sequences.

See also

What is the canonical sequence? http://www.uniprot.org/faq/30

How to retrieve sets of protein sequences? http://www.uniprot.org/faq/38

Elisabeth

[Pablo Pareja]

Thanks for your quick answer, I have a few remarks about this though. Why not including information about the isoform like name and sequence in the XML files? Besides that, I just downloaded the file uniprot_sprot.fasta and there is no entry for the isoform Q8IYH5-1 here either. Should it be somewhere else?

[Elisabeth Gasteiger]

Isoform names are in the xml: see e.g. <name>1</name> in your above example.

Sequences are not, I can check for you whether this has ever been discussed.

Q8IYH5-1 is the displayed (canonical) sequence of Q8IYH5, and its sequence is therefore identical to that of Q8IYH5. uniprot_sprot.fasta has an entry for all canonical isoforms, including Q8IYH5.