Now that all the data is available from a SPARQL endpoint, we can get at it from web pages via JavaScript and integrate it with visualization tools such as Jmol:


The visualization prototype has these features so far:

  • Choose a locus (HLA-A, HLA-B, ...) where the list is populated via SPARQL.
  • Choose from PDB structures with alleles from that locus and load it into Jmol.
  • Turn the HLA chain off, large, or small; likewise the peptide.
  • Turn contact points off, large or small.
  • Pick a peptide residue and highlight its contacts.
  • Turn variations on this allele off, large, or small; likewise all variations on this locus.
  • Make ad-hoc sparql queries and show the results in Jmol.
  • See each query used by the JavaScript code in a log; follow a link from each query to a SPARQL query form where the query can be run and the results seen.

source: jmol-viz in svn; see also bug #159.

running instance (not stable): HLA Visual Query

See also: ImmPort/Demo sketch