University of Glasgow

Roderic Page

Continuing my on-again off-again relationship with the Semantic Web, I stumbled across a cool approach to visualising the results of SPARQL queries.

The Semantic Web made fun: d3sparql

![RDF Resource Description Framework
Icon](http://www.w3.org/RDF/icons/rdf_w3c_icon.128){style="float:right;padding:5px;"
border="0"}  
Readers of this blog will know that I\'m [sceptical about the current
value of linked data and RDF in biodiversity
informatics](http://iphylo.blogspot.co.uk/2011/10/final-thoughts-on-tdwg-rdf-challenge.html).
But I came across an interesting paper on RDF and biocuration that
suggests a good \"use case\" for RDF in constructing and curating
taxonomic databases.  
  
The paper is \"Catching inconsistencies with the semantic web: a
biocuration case study\" ([PDF
here](http://ceur-ws.org/Vol-952/paper_2.pdf)) by Jerven Bolleman and
Sebastien Gehant. The basic idea is that errors in databases (in this
case, [UniProt](http://www.uniprot.org)) can be flagged by constructing
queries in SPARQL that return results if there is a problem (for example
if a sequence annotation is contradictory).  
  
In recent posts I\'ve been [complaining about errors in the GBIF
taxonomy](http://iphylo.blogspot.co.uk/2013/06/more-gbif-fail.html),
notably duplicate taxa that are synonyms. One way to tackle this would
be to develop a set of SPARQL queries that we could use to flag
potential problems. For example, if two names are objective synonyms
then only one of them should be a node in the GBIF classification. If
both exist then we have a problem. If we know a name is a homonym of an
older name, but that name exists in the GBIF classification, then we
could flag that as an issue. We could also construct queries that flag
possible problems, even if we don\'t have precise information on
synonymy. For example, in [this
post](http://iphylo.blogspot.co.uk/2013/06/more-gbif-fail.html) I noted
that several frog species appear twice in the GBIF classification
because GBIF has aggregated classifications that put these frogs in
different genera. We could catch such cases by constructing a query to
check whether the same species name (specific epithet) appeared in
different genera within the same family.  
  
The advantage of using RDF and SPARQL in this context is that that the
queries are portable. Assuming everyone uses the same vocabulary (e.g.,
[the TDWG LSID
vocabularies](http://wiki.tdwg.org/twiki/bin/view/TAG/LsidVocs)) then
queries can be constructed by one person (e.g., me) and then used by
anyone who has their data in a triple store. We could develop a set of
\"taxonomy tests\" that anyone could apply to their database.  
  
This idea needs some more work, but it would be fun to play with some
data and see how many kinds of errors or issues we can catch in this
way.


Readers of this blog will know that I'm sceptical about the current value of linked data and RDF in biodiversity informatics. But I came across an interesting paper on RDF and biocuration that suggests a good "use case" for RDF in constructing and curating taxonomic databases.The paper is "Catching inconsistencies with the semantic web: a biocuration case study" (PDF here) by Jerven Bolleman and Sebastien Gehant.

A use case for RDF in taxonomy

[![](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgoQSJ-NUcQu4tE5Ux7U0b5p_UkkkhV0fbYFfayCd2gTeD3biMyNQPce1fJD2j6Y1UQaYz5mITfeA1xEVhxI1esvMYBwDAuPM8nVOHG1c10TqZW57F-EV4iqei1l3hNRH08UY9a/s320/taxo_splash.gif){#BLOGGER_PHOTO_ID_5356793237794613186
style="float:right; margin:0 0 10px 10px;cursor:pointer; cursor:hand;width: 320px; height: 288px;"
border="0"}](https://blogger.googleusercontent.com/img/b/R29vZ2xl/AVvXsEgoQSJ-NUcQu4tE5Ux7U0b5p_UkkkhV0fbYFfayCd2gTeD3biMyNQPce1fJD2j6Y1UQaYz5mITfeA1xEVhxI1esvMYBwDAuPM8nVOHG1c10TqZW57F-EV4iqei1l3hNRH08UY9a/s1600-h/taxo_splash.gif){onblur="try {parent.deselectBloggerImageGracefully();} catch(e) {}"}  
Lately I\'ve been returning to playing with RDF and triple stores. This
is a serious case of déjà vu, as two blogs I\'ve now abandoned will
testify ([bioGUID](http://bioguid.blogspot.com/) and
[SemAnt](http://semant.blogspot.com/)). Basically, a combination of
frustration with the tools, data cleaning, and the lack of identifiers
got in the way of making much progress. I gave up on triple stores for a
while, rolling my own [Entity--Attribute--Value (EAV)
database](http://iphylo.blogspot.com/2008/02/database-of-everything.html),
which I used for the [Elsevier
Challenge](http://iphylo.org/~rpage/challenge/www/) (EAV databases are
essentially key-value databases, [CouchDB](http://couchdb.apache.org/)
being a well-known example).  
  
Now, I\'m revisiting triple stores and SPARQL, partly because Linked
Data is gaining momentum, and partly because we now have a few LSID
providers, and [some decent
vocabularies](http://rs.tdwg.org/ontology/voc/TaxonName) from
[TDWG](http://www.tdwg.org). Having created a LSID resolver that [plays
nicely with Linked
Data](http://iphylo.blogspot.com/2009/04/lsids-http-uri-linked-data-and-bioguid.html)
(it also does the same thing for DOIs), it\'s time to dust off SPARQL
and see what can be done.  
  
One reason there\'s interest in having GUIDs and standard vocabularies
is so that we can link different sources of information together. But
more than just linking, we should be able to compute across these links
and learn new things, or at least add annotations from one database to
another.  
  
To make this concrete, take the NCBI [taxon
101855](http://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwtax.cgi?id=101855)
, *Lulworthia uniseptata*. If we visit the NCBI page we see links to
other resources, such as Index Fungorum [record
105488](http://www.indexfungorum.org/Names/namesrecord.asp?RecordId=105488),
which tells us that *Lulworthia uniseptata* was published in *Trans.
Mycol. Soc. Japan* 25(4): 382 (1984), and that the current name is
[*Lulwoana
uniseptata*](http://www.indexfungorum.org/Names/SynSpecies.asp?RecordID=369395),
which was published in *Mycol. Res.* 109(5): 562 (2005).  
  
Wouldn\'t it be nice to be able to automatically link these things
together? And wouldn\'t it be nice to have identifiers for the
literature, rather than only human-readable text strings? Using bioGUID,
we can discover that [*Mycol. Res.* 109(5): 562
(2005)](http://bioguid.info/openurl/?genre=article&title=Mycol.+Res.&volume=109&spage=&epage=&pages=562&date=)
has the DOI
[doi:10.1017/S0953756205002716](http://dx.doi.org/10.1017/S0953756205002716)
\-- I haven\'t found *Trans. Mycol. Soc. Japan* 25(4): 382 (1984) online
anywhere.  
  
Now, given that we have LSIDs for Index Fungorum, I can resolve
[urn:lsid:indexfungorum.org:names:369395](http://bioguid.info/urn:lsid:indexfungorum.org:names:369395)
and discover that  
  
urn:lsid:indexfungorum.org:names:369395 tname:hasBasionym
urn:lsid:indexfungorum.org:names:105488  
  
and, I can add the statement  
  
urn:lsid:indexfungorum.org:names:36939 tcommon:publishedInCitation
doi:10.1017/S0953756205002716  
  
What I\'d like to do is link this to the NCBI taxon, so that I can
display this additional knowledge in one place (i.e., there is an
additional name for this fungus, and where it is published). To do this,
I need the NCBI taxonomy in RDF. Turns out that [everyone and their
dog](http://www.google.com/search?q=ncbi+taxonomy+rdf) has been
generating RDF versions of the NCBI taxonomy, including
[Uniport](http://www.uniprot.org/taxonomy/) (source of the diagram
above). The problem is, each effort creates their own project-specific
vocabulary. For example , here is the record for NCBI taxon 101855 in
Uniprot RDF (<http://www.uniprot.org/taxonomy/101855>):  
  

``` brush:xml

    xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#" xmlns:owl="http://www.w3.org/2002/07/owl#"
    xmlns:dc="http://purl.org/dc/elements/1.1/">
    
        
        
        Lulworthia uniseptata
        Zalerion maritimum
        
        false
    
```

  
  
Uniprot has it\'s own vocabulary, <http://purl.uniprot.org/core/>. So,
what I\'d like to do is create a version of the NCBI taxonomy using
TDWG\'s [TaxonConcept](http://rs.tdwg.org/ontology/voc/TaxonConcept#)
vocabulary, so that it becomes straightforward to link NCBI to name
databases such as Index Fungorum, IPNI, Zoobank, and ION that are
serving taxon names.


Lately I've been returning to playing with RDF and triple stores. This is a serious case of déjà vu, as two blogs I've now abandoned will testify (bioGUID and SemAnt).  Basically, a combination of frustration with the tools, data cleaning, and the lack of identifiers got in the way of making much progress.

Rogue Scholar Beiträge

The Semantic Web made fun: d3sparql

A use case for RDF in taxonomy

NCBI taxonomy, TDWG vocabularies, and RDF