[Cdk-user] UniversalIsomorphismTester differences between CDK 1.0.3 and 1.4.4

SourceForge Headquarters 1320 Columbia Street Suite 310 San Diego, CA 92101 +1 (858) 422-6466

We are experiencing different results using the UniversalIsomorphismTester between CDK versions 1.0.3 and 1.4.4, particularly with aromatic groups. If the test code is correct then it is possible that version 1.4.4 is wrong. The test code and output is below.

                public static void main(String[] args) {
            String[] fgSmiles = { "O=C1C=CC=CC1=O", "O=C1C=CC(=O)C=C1", "ClC=N" };
            // SIDs 26755257, 29215022
            String[] testSmiles = {
                        "CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)CO)O)[NH3+])O",
                        "CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)C)O)N)O",
                        "CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)C)O)N)O.Cl",
                        "CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)CO)O)N)O.Cl",
                        "CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)C)O)N)O",
                        "C1CC2=CC=CC=C2C1NC3=C(C(=O)C4=C(C3=O)N=CC=C4)Cl",
                        "CC1(OCC(O1)CSC2=C(C(=O)C3=C(C2=O)C=CC=N3)SCC4COC(O4)(C)C)C",
                        "CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)CO)O)N)O",
                        "CC1COC2=C1C(=O)C(=O)C3=C2C=CC4=C3CCCC4(C)C",
                        "C1=CC=C2C(=C1)C=CC3=C2C=CC4=C3C(=O)C=CC4=O",
                        "CC1=CC(=O)C2=C(C1=O)C=CC=C2O",
                        "CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)C)O)N)O",
                        "C1=CC=C2C(=C1)C(=O)C3=C(C2=O)C(=C(C=C3NC4=CC(=C(C=C4)NC5=NC(=NC(=N5)NC6=CC=C(C=C6)S(=O)(=O)[O-])Cl)S(=O)(=O)[O-])S(=O)(=O)[O-])N.[Na+].[Na+].[Na+]",
                        "CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5O)O)(C(=O)C)O)N)O",
                        "C1=CC=C2C(=C1)C(=O)C3=C(C2=O)C(=C(C=C3NC4=CC(=C(C=C4)NC5=NC(=NC(=N5)NC6=CC=C(C=C6)S(=O)(=O)[O-])Cl)S(=O)(=O)[O-])S(=O)(=O)[O-])N",
                        "C1=CC(=C(C=C1CN=C(N)NC(=O)C2=C(N=C(C(=N2)Cl)N)N)Cl)Cl.Cl", "C1=CC2=C(C(=O)C=CC2=O)C(=C1)O" };
            try {

                  System.out.print("Smiles\t");
                  for (String group : fgSmiles) {
                        System.out.print(group + "\t");
                  }
                  System.out.println();
                  for (String smiles : testSmiles) {
                        System.out.print(smiles);
                        IMolecule mol1 = getMolecule(smiles);
                        for (String ss : fgSmiles) {
                              IMolecule mol2 = getMolecule(ss);
                              System.out.print("\t" + UniversalIsomorphismTester.isSubgraph(mol1, mol2) );
                        }
                        System.out.println();
                  }

            } catch (Exception e) {
                  // TODO Auto-generated catch block
                  e.printStackTrace();
            }
}

/* CDK Version 1.4.4 */
      public static IMolecule getMolecule(String structure) throws CDKException, IOException {
            IMolecule mol = null;
            IChemObjectBuilder builder = DefaultChemObjectBuilder.getInstance();

            SMILESReader sr = new SMILESReader(new StringReader(structure));
            MoleculeSet set = (MoleculeSet) sr.read(builder.newInstance(new MoleculeSet().getClass()));
            mol = set.getMolecule(0);

            CDKHydrogenAdder adder = CDKHydrogenAdder.getInstance(builder);
            adder.addImplicitHydrogens(mol);

            return mol;
      }

      /* CDK Version 1.0.3 */
      public static IMolecule getMolecule(String structure) throws CDKException, IOException {
            IMolecule mol = null;
            DefaultChemObjectBuilder builder = DefaultChemObjectBuilder.getInstance();
            SMILESReader sr = new SMILESReader(new StringReader(structure));
            MoleculeSet set = (MoleculeSet) sr.read(builder.newMoleculeSet());
            mol = set.getMolecule(0);

            HydrogenAdder adder = new HydrogenAdder();
            adder.addImplicitHydrogensToSatisfyValency(mol);

            return mol;
      }

CDK 1.4.4

CDK 1.0.3

CDK 1.4.4

CDK 1.0.3

CDK 1.4.4

CDK 1.0.3

Smiles

O=C1C=CC(=O)C=C1

O=C1C=CC(=O)C=C1

O=C1C=CC=CC1=O

O=C1C=CC=CC1=O

ClC=N

ClC=N

CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)CO)O)[NH3+])O

FALSE

TRUE

CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)C)O)N)O

FALSE

TRUE

CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)C)O)N)O.Cl

FALSE

TRUE

CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)CO)O)N)O.Cl

FALSE

TRUE

CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)C)O)N)O

FALSE

TRUE

C1CC2=CC=CC=C2C1NC3=C(C(=O)C4=C(C3=O)N=CC=C4)Cl

FALSE

TRUE

CC1(OCC(O1)CSC2=C(C(=O)C3=C(C2=O)C=CC=N3)SCC4COC(O4)(C)C)C

FALSE

TRUE

CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)CO)O)N)O

FALSE

TRUE

C1=CC=C2C(=C1)C=CC3=C2C=CC4=C3C(=O)C=CC4=O

FALSE

TRUE

CC1=CC(=O)C2=C(C1=O)C=CC=C2O

FALSE

TRUE

CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5OC)O)(C(=O)C)O)N)O

FALSE

TRUE

CC1C(C(CC(O1)OC2CC(CC3=C(C4=C(C(=C23)O)C(=O)C5=C(C4=O)C=CC=C5O)O)(C(=O)C)O)N)O

FALSE

TRUE

C1=CC2=C(C(=O)C=CC2=O)C(=C1)O

FALSE

TRUE

CC1COC2=C1C(=O)C(=O)C3=C2C=CC4=C3CCCC4(C)C

FALSE

TRUE

C1=CC=C2C(=C1)C(=O)C3=C(C2=O)C(=C(C=C3NC4=CC(=C(C=C4)NC5=NC(=NC(=N5)NC6=CC=C(C=C6)S(=O)(=O)[O-])Cl)S(=O)(=O)[O-])S(=O)(=O)[O-])N.[Na+].[Na+].[Na+]

FALSE

TRUE

C1=CC=C2C(=C1)C(=O)C3=C(C2=O)C(=C(C=C3NC4=CC(=C(C=C4)NC5=NC(=NC(=N5)NC6=CC=C(C=C6)S(=O)(=O)[O-])Cl)S(=O)(=O)[O-])S(=O)(=O)[O-])N

FALSE

TRUE

C1=CC(=C(C=C1CN=C(N)NC(=O)C2=C(N=C(C(=N2)Cl)N)N)Cl)Cl.Cl

FALSE

TRUE