How Easy Is It To Make A Sparql Endpoint?

Question

[Andra Waagmeester]

I saw this nice overview on SPARQL endpoints. I was wondering how complete it was regarding biological resources but more importantly how can I make my resource such an endpoint. I googled on the topic, but could only find tutorials in making sparql clients, not sparql engines. I am aware that I should actually ask this question on stackoverflow (which I might still do), but I would appreciate the answer from a bioinformaticians point of view.

Answer 1

[Egon Willighagen]

There are a couple of options, including setting up a local store, using e.g. Virtuoso (which comes as package with Debian and Ubuntu). Alternatively, you use a hosting service, e.g. for Open Data offered by Science3.0, and Talis. More details about this at SemanticOverflow.

A third option is particularly interesting for small data sets: use an externals SPARQL service (like sparql.org), such as I used in this blog post, or more accurately used indirectly by using RDFaDev.

[Andra Waagmeester]

There is a semanticoverflow. NICE!

Answer 2

[Joachim]

BioMart 0.8rc6 (http://www.biomart.org/) will have a RESTful SPARQL-API build in.

There is experimental support for SPARQL in BioMart available right now, but with rc6 the implementation will be officially supported. The RESTful API will also allow you to retrieve an ontology-view of the mart, which is automatically generated.

Feel free to subscribe to the BioMart mailing-list (http://www.biomart.org/contact.html) to get the latest info when rc6 is released next month.

[Egon Willighagen]

Wow! BioMart++!

[Pierre Lindenbaum]

+1 Wow! BioMart++!

Answer 3

[Pierre Lindenbaum]

• at Biohackathon2010 we learned how to setup a RDF datastore using virtuoso. See http://hackathon3.dbcls.jp/wiki/VirutuosoQuickPrimer (the site is down due to electricity shortage resulting from the earthquake (BTW, I hope everybody is safe there) ) . The page is available in the google-cache, search for site:hackathon3.dbcls.jp VirutuosoQuickPrimer
• we also learned how to use D2RQ to map an existing RDBMS database to a RDF store. See my blog post.

[Egon Willighagen]

Ouch... "Suspension of a part of the LSDB/DBCLS services to save electricity" BTW... lovely "Virutuoso" typo :)

Answer 4

[Samuel Lampa]

The ARC PHP RDF library includes a SPARQL endpoint, and is a really easy way to set up a SPARQL endpoint, and should not require more software stack than the standard XAMP (OS, Apache, MySQL and PHP) stack supported by most hosting providers.

There is also ARC based solutions for exposing the RDF data of common CMS:es in a SPARQL endpoint, which could be even easier, depending on your use case. See e.g:

• SPARQL for Drupal
• RDFIO for Semantic MediaWiki

Answer 5

[Andra Waagmeester]

I am currently using 4store.org as a sparql endpoint. This proved to be very easy to use and install.