HCLS/HCLS semantic web map graphviz code

From W3C Wiki
Jump to: navigation, search

Graphviz code


digraph G {	
edge [dir="none"];

subgraph cluster_molecular_interactions {
  label = "molecular interactions";
  inoh_biopax; // 1        http://www.inoh.org/ http://bio2rdf.org/inoh:Identifier
  reactome_biopax; // 2    http://reactome.org/ http://bio2rdf.org/reactome:Identifier
  pdsp_kidb; // 1          http://esw.w3.org/topic/HCLSIG_BioRDF_Subgroup/Data/KI_database
  neurocommons; // 1       http://sw.neurocommons.org/2007/text-mining.html
}

subgraph cluster_small_molecules {
  label = "small molecules";
  chebi; // 2              http://www.obofoundry.org/cgi-bin/detail.cgi?id=chebi http://bio2rdf.org/chebi:Identifier
  pubchem; // 1            http://bio2rdf.org/pubchem:Identifier
}

subgraph cluster_proteins {
  label = "proteins"
  uniprot_rdf; // 2        http://beta.uniprot.org/ http://bio2rdf.org/uniprot:Identifier http://bio2rdf.org/uniparc:Identifier http://bio2rdf.org/uniref:Identifier
  uniprot_purl // 1
  antibodies; // ???
}

subgraph cluster_cells {
  label = "cells";
  cell_ontology; // 1
  ccdb_sao; // 2
  neuron_db; // 2
  go_cellular_location; //3 http://bio2rdf.org/go:Identifier
}

subgraph cluster_biological_investigation {
  label = "biological investigation";
  obi; // 2
  expo; // 1
  evidence_codes; // 1
  mged;  //   microarray data
}

subgraph cluster_anatomy {
  label = "anatomy";
  bams; // 2
  birnlex; // 2
  caro; // 2
  fma_lite; // ???
  allen_brain_atlas; // 2   proximity to "genes"
}

subgraph cluster_diseases {
  label = "diseases";
  swan; // 1
  disease_ontology; // 2
  brainpharm; // 1
}

subgraph cluster_genes {
  label = "genes";
  sequence_ontology; // 1
  entrez_gene; // 1 http://bio2rdf.org/geneid:Identifier
  alzgene; // 1    Where is the RDF/OWL ?
  homologene; // 1 http://bio2rdf.org/homologene:Identifier
}

subgraph cluster_physiology {
  label = "physiology";
  go_biological_process; // 3 http://bio2rdf.org/go:Identifier
}

subgraph cluster_phenotypes {
  label = "phenotypes";
  pato; // 2
  mammalian_phenotype; //   proximity to genes and diseases
}

subgraph cluster_clinical_investigation {
  label = "clinical investigation and patient care";
  oci;  // 1
  pomr; // 2
  galen; // 3     http://www.co-ode.org/galen/
  acpp; // 1
  // RIM ontology;  // http://veggente.berlios.de/ns/RIMOntology
  // some ontologies on http://bioinfo.icapture.ubc.ca/ might also be of interest
}

subgraph cluster_bibliography {
  pubmed_purl; // 1  ??? http://bio2rdf.org/pubmed:Identifier (not globally available due to copyright restrictions but available for personal use on request)
  pubmed_lsid; // 1
}

// still unsorted ontologies

go_molecular_function; // http://bio2rdf.org/go:Identifier

mesh; // http://bio2rdf.org/mesh:Identifier

// connections

bams -> caro;
bams -> neuron_db;
brainpharm -> neuron_db;
birnlex -> ccdb_sao;
birnlex -> neuron_db;
ccdb_sao -> neuron_db;
chebi -> pdsp_kidb;
galen -> pomr;
go_molecular_function -> neuron_db;
neuron_db -> pdsp_kidb;
neuron_db -> pubmed_purl;
neuron_db -> uniprot_purl;
pdsp_kidb -> pubmed_purl;
pdsp_kidb -> uniprot_purl;


}


Syntax and content guidelines

Example output of Graphviz (fdp.exe):

[1]

(This image file might be out of date and might not represent the newest version of the graphviz code above)

This Graphviz layout will be used as a blueprint for drawing the HCLS Semantic Web map

Further Notes

Eric Jain writes:

1. UniProt is not one database, but several (some of which may be "duplicated" elsewhere in your map, and some are "duplicates" themselves).

It may therefore be worth adding these relationships to your map:

UniProtKB (UniProt) -- UniRef (UniProt) UniProtKB (UniProt) -- UniParc (UniProt) UniProtKB (UniProt) -- Taxonomy (UniProt) UniProtKB (UniProt) -- Keywords (UniProt) UniProtKB (UniProt) -- Pathways (UniProt) UniProtKB (UniProt) -- GO (UniProt) Keywords (UniProt) -- GO (UniProt) UniRef (UniProt) -- Taxonomy (UniProt) UniParc (UniProt) -- Taxonomy (UniProt) UniProtKB (UniProt) -- Citations (UniProt) UniProtKB (UniProt) -- Enzyme (UniProt)

2. I've created some mappings between the UniProt RDF distribution and several RDFified databases hosted by bio2rdf:

<http://dev.isb-sib.ch/projects/uniprot-rdf/mappings/>

This would allow us to add the following connections:

UniProtKB (UniProt) -- OMIM (bio2rdf) UniProtKB (UniProt) -- PDB (bio2rdf) UniProtKB (UniProt) -- MGI (bio2rdf) UniProtKB (UniProt) -- GenBank (bio2rdf)

Also, these, but they seem to be broken on the bio2rdf server right now:

Citations (UniProt) -- PubMed (bio2rdf) UniProtKB (UniProt) -- InterPro (bio2rdf) UniProtKB (UniProt) -- Ensembl (bio2rdf)

Comment by PeterAnsell: Except for PubMed, Uniparc and Uniref, the Bio2RDF endpoints should be functioning correctly. You may need to use the sourceforge bio2rdf package to get information in the normalised Bio2RDF format though. Bio2RDF mirror all of the UniProtKB under the same namespaces, ie, uniparc, uniref, citations, keywords, enzyme.