bio2rdf / bio2rdf-rest-talend Goto Github PK

Currently there is no support for entity resolution for arbitrary entities that appear in Bio2RDF datasets. Consider the statistics page for affymetrix probesets
http://download.bio2rdf.org/release/3/html/affymetrix.html

The first type http://bio2rdf.org/wormbase_vocabulary:Resource does not resolve even though WormBase is part of the Bio2RDF distribution.

The second type http://bio2rdf.org/uniprot_vocabulary:Resource does not resolve.

the web application should send out a federated query to SPARQL endpoints to gather triples to describe these terms.

@vemonet

Git is designed to help with versioning of text files. Publishing the essential code as ZIP files makes it impossible to practically trace bugs and bug fixes in Git.

I propose that we include the provenance for a REST call in the returned results. I suggest we offer quad-based nanopublications [1] with the resulting triples in an "assertion graph" and the provenance of how and where we got those triples in the "provenance graph". here is a quick mockeup:

 @prefix : <http://nanopublications.bio2rdf.org/> .
 @prefix np:  <http://www.nanopub.org/nschema#> .
 @prefix prov: <http://www.w3.org/ns/prov#> .
 @prefix xsd:  <http://www.w3.org/2001/XMLSchema#> .

 # nanopublication
 :np-uuid {
   np:hasPubInfo :np-pubinfo-uuid;
   np:hasAssertion :np-assertion-uuid;
   np:hasProvenance :np-provenance-uuid;
  }

 #  nanopublication metadata
 :np-pubinfo-uuid {
  dct:title "Bio2RDF SPARQL query nanopublication"; 
  prov:generatedAtTime  "YYYY-MM-DDTHH:MM:SSZ"^^xsd:dateTime;
  dct:publisher <http://bio2rdf.org>;
  dct:license <http://creativecommons.org/licenses/by/4.0/>;
  cito:citesAsAuthority {} . # bio2rdf publications here
 }

 # nanopublication assertion
 :np-assertion-uuid {
    # query triples here
  }

  # nanopublication provenance
 :np-provenance-uuid {
   :np-assertion-uuid pav:wasGeneratedBy :np-agent-uuid .
   :np-agent-uuid
      a :web-service; # find appropriate type 
      prov:startedAtTime "";
      prov:endedAtTime "";
      prov:specializationOf <http://bio2rdf.org>;
      prov:used bio2rdf:dataset1, bio2rdf:dataset2, ...  # versioned dataset graph uris
      prov:used :url_query_1, :url_query_2, ... # full URL of SPARQL queries for reproducibility.
 }

[1] http://www.nanopub.org/2013/WD-guidelines-20131215/

It looks like whatever libraries are being used to handle the content-negotiation at the Bio2RDF end are not dealing with these complex (but valid!) request headers.

curl -H 'Accept: application/rdf+xml, application/xml; q=0.8, text/xml; q=0.7, application/rss+xml; q=0.3, /; q=0.2' --dump-header -http://bio2rdf.org/hgnc_vocabulary:approved-name
HTTP/1.1 404 Not Found
Date: Fri, 13 Jun 2014 08:02:39 GMT
Server: Jetty(8.1.12.v20130726)
Content-Length: 0

$ curl -H 'Accept: application/rdf+xml' --dump-header -http://bio2rdf.org/hgnc_vocabulary:approved-name
HTTP/1.1 200 OK
Date: Fri, 13 Jun 2014 08:03:11 GMT
Server: Jetty(8.1.12.v20130726)
Content-Type: application/rdf+xml
Content-Length: 1720

@vemonet

@vemonet @ansell
Should we consider implementing the Linked Data API (https://code.google.com/p/linked-data-api/wiki/Specification)?

• logo

• tagline is "linked data for the life sciences"

• auto-completion search bar

• university logos made smaller, and positioned at the bottom of the page.

• links to wiki pages for

  a) about - description, people, organizations, publications

  b) contact - mailing list, admin email

  c) sources - datasets (description + example identifiers), sparql endpoints, facet search interfaces

  d) SPARQL - list of example queries

(all other things removed)

@vemonet

The vast majority of the datasets in Bio2RDF are not licensed under CC-BY, but the DataHub upload script 0f39015 has hardcoded that they are all CC-BY licensed, and that they are all "open" in the OpenDefinition.org sense, which they are most certainly not if you look into their licensing conditions one by one.

Please do something intelligent to map the actual licenses to the DataHub scheme, defaulting to "not-open" if the license is unknown or the license is known by it does not fulfill all of the OpenDefinition.org points.

@vemonet
In order to track changes on commits, and to make inline changes on the code, the codebase should not be in a zip file.

We should should have a document that specifies the API functionality and include their SPARQL queries. Please create a wiki page to keep this documentation.
@vemonet

It would be invaluable to have a service (e.g. /endpoint-status/[all|name]) to check the "status" of the endpoints for which we depend on. The service should send a SPARQL query like "describe http://bio2rdf.org/rest-service/ping" and get an empty turtle document.

@vemonet

You can get more information about http://bio2rdf.org/ns_vocabulary:Resource entities from the Life Science Registry. For instance, the :Resource entity defined by the sgd_vocabulary is linked to the LSR entry :

http://lsr.bio2rdf.org/describe/?url=http%3A%2F%2Fbio2rdf.org%2Fsgd_vocabulary:Resource

@vemonet

http://linkeddata.uriburner.com/about/html/http://bio2rdf.org/drugbank:DB00363

not sure what the problem is...

Please list the set of format types that are supported in the API on their example pages

For instance, on http://bio2rdf.org/test/affymetrix

in -> http://bio2rdf.org/affymetrix/describe/rdf/affymetrix:185061_at
there is "rdf", but really this corresponds to "rdf+xml"
and i apparent can also get "nt" for ntriples, and 'n3' or 'turtle'
what other formats are supported?

It looks like there is some rewriting of the content because the rdf:type should be
http://bio2rdf.org/hgnc_vocabulary:Resource

but is being incorrectly rewritten to
http://hgnc_vocabulary.bio2rdf.org/resource

$ curl -H 'Accept: application/rdf+xml' http://bio2rdf.org/hgnc_vocabulary:approved-name
HTTP/1.1 200 OK
Date: Fri, 13 Jun 2014 08:03:11 GMT
Server: Jetty(8.1.12.v20130726)
Content-Type: application/rdf+xml
Content-Length: 1720

<rdf:RDF
    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:j.0="http://bio2rdf.org/hgnc_vocabulary:"
    xmlns:owl="http://www.w3.org/2002/07/owl#"
    xmlns:dc="http://purl.org/dc/elements/1.1/"
    xmlns:dcterms="http://purl.org/dc/terms/"
    xmlns:j.1="http://hgnc_vocabulary.bio2rdf.org/"
    xmlns:void="http://rdfs.org/ns/void#"
    xmlns:ns5="http://bio2rdf.org/bio2rdf_vocabulary:"
    xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"  >
  <rdf:Description rdf:about="http://bio2rdf.org/hgnc_vocabulary:approved-name">
    <rdf:type rdf:resource="http://hgnc_vocabulary.bio2rdf.org/resource"/>
    <ns5:uri rdf:datatype="http://www.w3.org/2001/XMLSchema#string">http://bio2rdf.org/hgnc_vocabulary:approved-name</ns5:uri>
    <dc:identifier>hgnc_vocabulary:approved-name</dc:identifier>
    <rdf:type rdf:resource="http://www.w3.org/2002/07/owl#DatatypeProperty"/>
    <ns5:namespace rdf:datatype="http://www.w3.org/2001/XMLSchema#string">hgnc_vocabulary</ns5:namespace>
    <rdfs:label xml:lang="en">approved name [hgnc_vocabulary:approved-name]</rdfs:label>
    <rdfs:label rdf:datatype="http://www.w3.org/2001/XMLSchema#string">approved name [hgnc_vocabulary:approved-name]</rdfs:label>
    <dcterms:title xml:lang="en">approved name</dcterms:title>
    <rdf:type rdf:resource="http://bio2rdf.org/hgnc_vocabulary:Resource"/>
    <ns5:identifier rdf:datatype="http://www.w3.org/2001/XMLSchema#string">approved-name</ns5:identifier>
    <dcterms:identifier rdf:datatype="http://www.w3.org/2001/XMLSchema#string">hgnc_vocabulary:approved-name</dcterms:identifier>
    <void:inDataset rdf:resource="http://bio2rdf.org/hgnc_resource:bio2rdf.dataset.hgnc.R3"/>
  </rdf:Description>
</rdf:RDF>