I wrote several SPARQL queries to extract biological/chemical networks from Bio2RDF and LODD. Same method can be applied to Chem2Bio2RDF too.

Case 1:
Glycan reaction network (From KEGG reaction, glycan and pathway).

PREFIX rn: <http://bio2rdf.org/rn:>
PREFIX gl: <http://bio2rdf.org/gl:>
PREFIX kegg: <http://bio2rdf.org/ns/kegg#>
PREFIX bio2rdf: <http://bio2rdf.org/ns/bio2rdf#>
PREFIX dc: <http://purl.org/dc/elements/1.1/>
 
SELECT ?substrate ?product ?reaction ?pathway
FROM <http://bio2rdf.org/kegg/reaction>
FROM NAMED <http://bio2rdf.org/kegg/glycan>
WHERE {
?reaction kegg:xProduct ?product .
?reaction kegg:xSubstrate ?substrate .
GRAPH <http://bio2rdf.org/kegg/glycan> {
?substrate a kegg:Glycan .
?product a kegg:Glycan .
}
?reaction kegg:xPathway ?pathway
}
 
GROUP BY ?substrate ?product



Separate query is used to extract the pathway title, details about the glycan, and reactions from the RDF graphs. The graph is created in an XGMML format, as attached here.

Here is the extracted network (from Cytoscape)

glycan.jpg

I also extracted drug-target network, and a protein-protein interaction network.

Cast 2: Drug-Target Network from DrugBank RDF graph of LODD
drug-target.jpg
Visualizing details of drugs in Drug-Target Network using ChemVIZ
chemviz.jpg