/uberlibsbgn

A simpler, more complete, less verbose, dev-friendly version of libsbgn

Primary LanguageJava

uberlibsbgn

API goal

// simplest building possible
Macromolecule m = new Macromolecule("PROT") // don't care about id
  .position(10,10)                          // don't care about sizes
  .addUnitOfInfo("receptor", TOP)           // don't care about computing auxiliary bboxes              
  .multimer();                              // don't care about class name management
  
// simple checks
m.hasUnitOfInfo(regexp);
m.isIn(complexOrCompartment);
m.getOutgoingArcs();
m.getDirectNeighbors();

// simple composite management
complex.include(m);
complex.getSubunits();
compartment.getIncluded();

// easy and efficient indexing
map.getAll("macromolecule");
map.getTopLevelGlyphs();

// easy traversing
map.visitRecursive();
map.visitEPNs();
map.visitTopLevel();

TODO

  • finish index manager rework
  • ports
  • get rid of fluent glyph setters, impossible to keep it consistent
  • start notes and extension
  • style management (find good lib for color and gradient management, javafx should do)
  • RDF (with rdf4j probably)
  • interface with some graph lib (jgraphT probably)
  • use jsoup for notes management and possibly extensions
  • refactor glyph factory/builder
  • more complete id management allow ids from sbgnml for arcs and aux units, maybe other entities
  • and so much more...

Ideas

  • canonicalize sbgnml with org.apache.xml.security from santuario then diff possible with diff libs
  • possible to convert whole xml to rdf directly using rdf mapper (carml)
  • possible to mine pathway informations directly from any rdf, so building sbgn from any detailed enough knowledgebase should be feasible