Nadia Anwar

2008-11-12T11:10:00.000-08:00

University Blog Page

2007-09-25T04:01:00.001-07:00

I have installed word-press on my server at Uni.

This is the link. RasorBlog

2007-07-09T07:41:00.001-07:00

Answer:
You can't have two tables in one model!
After creating a table you then create a model - this fails since the model already exists and as far as I can tell there is no SDO_RDF.add_to_model blah blah.

This may not even matter since all the models are stored in the same network and the tables store references to the RDF data that are loaded into the network.

So with better understanding: now the question is, one table/model for each RDF data file or throw everything together into one enormous table/model?? From the query examples I think I want to have all my data in one model as per

select x, y, name from
TABLE(SDO_RDF_MATCH(
‘(:Tom :hasParent ?y)
(?y :hasFather ?x)
(?x :name ?name)’,
SDO_RDF_Models(‘family'), ...);

Need to figure out if SDO_RDF_Models(‘family','order','plan') is possible???

2007-07-09T06:35:00.000-07:00

One table or two?
I have been figuring out the RDF capability in Oracle and I'm now stuck. I created a table rdf_dev, then created a model (as per the manual) and I have loaded one data set. I am stuck on what to do with the second data set. Do I create a new table (this seems non-sensicle since it will have the same structure and will be in the same model) or do I create a separate table for each data set in the model??? I have a feeling one table- one model is not going to scale very well. But - how is querying across these tables going to work? Do the queries run over the model or over the tables? Also, how do the inference rules run - again across the model or a across the table?? More reading.....

RDF serialization with SQLX

2007-06-21T05:35:00.002-07:00

Look how easy it is!!!

select XMLELEMENT("rdf:description",
XMLATTRIBUTES(assertion_id as "rdf:about=http://spira.zoology.gla.ac.uk/tcl_assertion.php?id"),
XMLELEMENT("assertion:source", get_source_name(source_id)),
XMLELEMENT("assertion:dbsourceid", dbsource_id),
XMLELEMENT("assertion:rank", rank),
XMLELEMENT("assertion:kingdom", kingdom),
XMLELEMENT("assertion:name",
XMLATTRIBUTES(get_name_text(name_id) as "rdf:resource=http://spira.zoology.gla.ac.uk/tcl_name.php?id"))
)
from assertion

TCl search WSDL

2007-06-13T07:41:00.000-07:00

I have built a SOAP service for the little useful PHP scripts I have built. The URL is
http://spira.zoology.gla.ac.uk/app/nusoap.
I have tested them all and they work as expected. There is a bug in one of the functions (getclassification_ncbi) which gets the classification string for NCBI so I haven't included it. The getclassification_itis is working. I haven't figured out why the NCBI classification fails it must be something in the data because the code is exactly the same.
Here is an example client code in PHP that calls the functions in the service

require_once('./lib/nusoap.php');
// Get inputs
$client = new soapclient('http://spira.zoology.gla.ac.uk/app/nusoap/tclsearchwsdl.php');
$int = array('int'=>'1818'); # AN ITIS or NCBI ID
$param = array('string'=>$_GET['string']);
$name_id = array('name_id'=>$_GET['name_id']);
// Call the functions
$result = $client->call('getName', $param);
$dot = $client->call('createDot', $param);
$itischildrendot = $client->call('getchildrendot_itis', $name_id);
$ncbichildrendot = $client->call('getchildrendot_ncbi', $name_id);
$classification = $client->call('getclassification', $param);
$name = $client->call('getNameByID', $int);
$ids = $client->call('getids', $param);
$assertions = $client->call('getassertions', $param);
$hierarchy = $client->call('gethierarchy', $param);
$synonym = $client->call('getsynonym', $param);
$vernacular = $client->call('getvernacular', $param);
// Display the results
print_r($result);
print_r($dot);
print_r($itischildrendot);
print_r($ncbichildrendot);
print_r($classification);
print_r($name);
print_r($ids);
print_r($assertions);
print_r($hierarchy);
print_r($synonym);
print_r($vernacular);
?>

I am still writing this chapter up, once it's done I'll copy more details on this here.

My other work......

2007-06-01T11:18:00.000-07:00

As part of the RASOR project, I have been tasked to implement technologies to interface across proteomic and genomic data. The focus of this project is improved data handling, storage and distribution through an integrated LIMS systems as a foundation to the establishment of an integrated relational database. On the face of it, this is classic data integration of two heterogeneous data systems, however given the nature of the data sources, proteomics data and genomics data, actual integration would be minimal since the overlapping data elements are few. The data are semantically different and therefore not only are they difficult to physically integrate but the process would add little value to the data itself. Since the actual reason for integration is to query the data as a unit, it is more important to the data users to have the data in a form that allows querying across these data. Semantic integration promises to provide exactly this capability.

The rasor project proposes to use semantic web technologies to interface across genomic and proteomic data. At the moment I am focusing on Globally Unique Identifers. URI's, Uniform Resource Identifier, forms the base technology on which RDF, Resource Description Framework was developed. All things on the Semantic Web are resources, and all resources may be identified by URIs. LSID's a type of URI (URN, Uniform Resource Name (URN): a type of URI that identifies a resource by name in a particular namespace) form one potential mechanism of globally unique identification.

PhD

2007-06-01T11:05:00.000-07:00

I am currently going through the painful process of writing up my PhD. I haven't decided on a title yet. The official title is "Tooling up for the Tree of Life", I have also been flaunting a very vague title "Phyloinformatics". Essentially, I built a data warehouse of taxonomic names for the purposes of of building a taxonomic backbone into TreeBASE. The warehouse contains names from multiple data sources so that there is broad coverage. The database, differs from most other taxonomic name servers in that the database also stores the individual classifications from each of the datasources, which can be used to perform hierarchical queries. The purpose of this database is to enable a kind of query expansion on taxon names. The best example is when someone submits a search using a vernacular name like birds, this name should translate (expand) to the latin name (Aves) and hierarchically to include all children of the term.