Hello, The term 'subgraph isomorphism' is (I think) usually used to mean finding a subgraph of a graph G that is isomorphic to another graph H. What you describe sounds like the more difficult problem of finding all common subgraphs of G and H.
So, I am not so sure that the javadocs are unclear in this case. There is the possibility that H may be found multiple times in G (a polymer, for example, will have multiple copies of the monomer in it), which is presumably why the return value of these methods are lists. As for how this could be done - well the simplest (and most expensive) way would be to enumerate every (connected?) subgraph of H, testing each one to see if it is also a subgraph of G. Gilleain Torrance On Mon, Mar 8, 2010 at 11:10 AM, Leonid Chepelev <[email protected]> wrote: > Hello All, > > When trying to find all the subgraph isomorphisms, one would anticipate from > the javadocs that the use of the > UniversalIsomorphismTester.getSubgraphAtomsMaps of CDK 1.2.5 would > return ALL the subgraph isomorphisms in a molecule of interest. However, > that is not the case. I am posting the code below to provide an example of > this behaviour: for the two molecules in question, we can see that there is > only one MCS from the MCSS code (as one would expect), but also only two > lists of RMaps from getSubgraphAtomsMaps, indicating only two subgraphs > identified. > > Do you believe it would be possible to get ALL the subgraphs, no matter how > small, into the getSubgraphAtomsMaps, as one would anticipate from the > description, or is there another way to do this within (or outside CDK)? > > Am I doing things correctly? > > If the behaviour of getSubgraphAtomsMaps is intentionally such that it only > returns n biggest subgraph matches, may I kindly suggest that the javadoc > description be changed? > > Regards, > > Leonid Chepelev > > *************************************************************** > import java.util.Iterator; > import java.util.List; > import org.openscience.cdk.io.SMILESWriter; > import java.io.StringWriter; > import org.openscience.cdk.DefaultChemObjectBuilder; > import org.openscience.cdk.interfaces.IAtomContainer; > import org.openscience.cdk.isomorphism.mcss.RMap; > import org.openscience.cdk.Molecule; > import org.openscience.cdk.smiles.SmilesParser; > import org.openscience.cdk.isomorphism.UniversalIsomorphismTester; > public class Main { > public static void main(String[] args) throws Exception { > SmilesParser smilesParser = new > SmilesParser(DefaultChemObjectBuilder.getInstance()); > String referenceMolecule = "C1CCCCC1CC(=O)OCCCC"; > String testmolecule = "C1CCCCC1CC(=O)O"; > IAtomContainer molecule = > smilesParser.parseSmiles(referenceMolecule); > IAtomContainer lookup = smilesParser.parseSmiles(testmolecule); > List allthefragments = > UniversalIsomorphismTester.getOverlaps(molecule, lookup); > for (Iterator<IAtomContainer> i = allthefragments.iterator(); > i.hasNext( ); ) { > IAtomContainer s = i.next( ); > try { > StringWriter funna = new StringWriter(); > SMILESWriter writemysmiles = new SMILESWriter(); > writemysmiles.setWriter(funna); > Molecule themoleculetowrite = new Molecule(s); > writemysmiles.write(themoleculetowrite); > writemysmiles.close(); > System.out.println(funna.toString()); > funna.close(); > } catch (Exception e) { > System.out.println(e.toString()); > } > } > List<List<RMap>> atommappingsad = > UniversalIsomorphismTester.getSubgraphAtomsMaps(molecule, lookup); > for (List<RMap> j:atommappingsad) { > for (RMap estro:j){ > try { > int firsta = estro.getId1(); > int seconda = estro.getId2(); > System.out.println("Atom " + firsta + " in G1 maps to > atom " + seconda + " in G2."); > } catch (Exception e) { > System.out.println(e.toString()); > } > } > System.out.println("End of mapped fragment..."); > } > } > } > ------------------------------------------------------------------------------ > Download Intel® Parallel Studio Eval > Try the new software tools for yourself. Speed compiling, find bugs > proactively, and fine-tune applications for parallel performance. > See why Intel Parallel Studio got high marks during beta. > http://p.sf.net/sfu/intel-sw-dev > _______________________________________________ > Cdk-user mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/cdk-user > > ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Cdk-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/cdk-user
