Does UniProt have complete SNP information (from dbSNP)?

Question

[ajingnk]

The recent UniProt website is fabulous, containing a complete collection of links to different resources. UniProt itself could be a very good resource for data mining for biological information.

I have a few questions regarding the SNP information in UniProt.

1. SNP information could be in two different section: Pathology and Biotech section and Sequence section. Is this from the annotation of disease SNP in dbSNP?
2. I want to identify disease SNP from natural variance. From xml file of UniProt (like: http://www.uniprot.org/uniprot/Q96CV9.xml), are all "feature" entries of variant with "evident" key disease related? Which is like following line: <feature type="sequence variant" description="In GLC1E." id="VAR_021537" evidence="16 17 18">
3. How does UniProt do the mapping between dbSNP and UniProt? Just by sequence alignment of mRNA? Is the mapping complete and correct? Or there is any heuristic behind.

Thanks

[me]

Thanks for liking the redesign!

Answer 1

[me]

1. The disease kind is explained in http://onlinelibrary.wiley.com/doi/10.1002/humu.22594/full you can read a bit more on the other kind here
2. This is a bit easier using RDF and sparql http://www.uniprot.org/uniprot/Q96CV9.rdf than in the XML format. See this example to get all natural variants with a xref and involved in disease entry.
3. Hope one of my other colleagues comes round to answer that ;)

[ajingnk]

Thanks! The SPARQL query on UniProt seems to be very powerful. Why do you need to define PREFIX rdfs and skos here? Is there any tutorial for SPARQL for UniProt?

[me]

Those two PREFIXes are just used a lot in different parts of the UniProt rdf model. This is a general intro to SPARQL using the UniProt/NCBI taxonomy data. But no tutorial currently exists :(