valence error error

In the following compound, Marvin should realize the N has a valence error, because it has a total electron count of 9 electrons, but it doesn't.

Code:

<?xml version="1.0" ?> <MDocument>   <MChemicalStruct>     <molecule molID="m1">       <atomArray           atomID="a1 a2 a3 a4 a5 a6"           elementType="C Br N H H H"           radical="0 monovalent monovalent 0 0 0"           x2="-8.285764546394347 -0.8745140957832334 -6.446805090904235 -5.805139036178589 -6.286388964653015 -3.976389021873474"           y2="6.70152818163236 6.701527943213781 7.048611198862394 8.92256868382295 4.495069866379103 6.516319637497267"           />       <bondArray>         <bond atomRefs2="a1 a3" order="1" />         <bond atomRefs2="a3 a4" order="1" />         <bond atomRefs2="a3 a5" order="1" />         <bond atomRefs2="a3 a6" order="1" />       </bondArray>     </molecule>   </MChemicalStruct> </MDocument>

Likewise, the C atom in this compound:

Code:

<?xml version="1.0" ?> <MDocument>   <MChemicalStruct>     <molecule molID="m1">       <atomArray           atomID="a1 a2 a3 a4 a5"           elementType="C H H H H"           formalCharge="-1 0 0 0 0"           radical="monovalent 0 0 0 0"           x2="-6.746597763299942 -5.233315241299588 -8.286597763299941 -7.145179092757823 -6.746597763299942"           y2="8.000590329368908 8.28620547008846 8.000590329368908 9.488116101854073 6.460590329368908"           />       <bondArray>         <bond atomRefs2="a1 a2" order="1" />         <bond atomRefs2="a1 a3" order="1" />         <bond atomRefs2="a1 a4" order="1" />         <bond atomRefs2="a1 a5" order="1" />       </bondArray>     </molecule>   </MChemicalStruct> </MDocument>

This error occurs in Marvin 5.1.1 as well as earlier versions back to at least Marvin 4.1.14.

We will check it.

Hi Bob,





We will fix this bug soon. Thank you again for the bug report.





Best regards,





Akos

This bug remains in JChem 5.3.1.

Yes, because it requires a significant change in the valence check algorithm, for which we had no capacity yet :(

We will check if a workaround is possible until that.

Best regards,

Akos

Oh, we've already had a workaround for years.  I was hoping it was no longer necessary.  Here's our code:

    public static final boolean INCLUDE_EXPLICIT_H = true;

    public static final boolean USE_BOND_ORDERS = false;



    public static void checkValence(Molecule mol) throws ValenceException {

        final String SELF = "ChemUtils.checkValence: ";

        // check for valence errors

        mol.valenceCheck();

        final String VALENCE_ERROR = "At least one of the atoms in your response "

                + "has an invalid valence.  If you don't see an atom underlined "

                + "in red, look for an atom that violates the octet rule, or "

                + "try expanding your shortcut groups.  ";

        for (int atomNum = 0; atomNum < mol.getAtomCount(); atomNum++) {

            final MolAtom atom = mol.getAtom(atomNum);

            if (atom.hasValenceError()) {

                throw new ValenceException(VALENCE_ERROR);

            } // if valence error

            // workaround for valence check bug for atoms with radical flag

            // such as 9-electron radicals

            int unsharedElecCt = mol.getLonePairCount(atomNum);

            final int radState = atom.getRadical();

            switch (radState) {

                case MolAtom.RAD1:

                    unsharedElecCt += 1; break;

                case MolAtom.RAD2:

                case MolAtom.RAD2_SINGLET:

                case MolAtom.RAD2_TRIPLET:

                    unsharedElecCt += 2; break;

                case MolAtom.RAD3:

                case MolAtom.RAD3_DOUBLET:

                case MolAtom.RAD3_QUARTET:

                    unsharedElecCt += 3; break;

                default: continue;

            } // radical state

            final int totalElecCt = unsharedElecCt

                    + atom.twicesumbonds(INCLUDE_EXPLICIT_H, USE_BOND_ORDERS)

                    + atom.getImplicitHcount() * 2;

            final int maxOuter = getMaxOuterElectrons(atom);

            if (totalElecCt > maxOuter) {

                throw new ValenceException(VALENCE_ERROR);

            } // if too many electrons

        } // for each atom

        // check for pentavalent N, as in nitro group

        final String PENTACOORD_N = "[$(*=,#[NX3]=,#*),$(*:n(=,#*):*),$(*=,#N#*)]";

        try {

            final Molecule pentacoordN = MolImporter.importMol(PENTACOORD_N);

            if (MolFunctions.containsSubstruct(mol, pentacoordN)) {

                throw new ValenceException(VALENCE_ERROR);

            } // if molecule contains pentavalent N

        } catch (MolFormatException e) { // unlikely

            ; // do nothing

        } // try

    } // checkValence(Molecule)



    public static int getMaxOuterElectrons(MolAtom atom) {

        return getMaxOuterElectrons(atom.getAtno());

    } // getMaxOuterElectrons(MolAtom)



    public static int getMaxOuterElectrons(int atno) {

        final int row = PeriodicSystem.getRow(atno);

        final int column = PeriodicSystem.getColumn(atno);

        if (row == 0) return 0; // unknown element

        else if (row == 1) return 2;

        else if (row == 2) return 8;

        else if (row == 3 || column >= 13) return 12; // hypervalent Si, S, etc.

        else if (row == 4 || row == 5 || column >= 4) return 18;

        else return 32;

    } // getMaxOuterElectrons(int)

MolFunctions.containsSubstruct() does a standard JChem substructure search.

Thanks for sharing your workaround.

Hi Bob,

We also found a workaround (but thanks for yours again), but we cannot push this fix into 5.3.2, because it is already in testing phase. The patch release coming next month will contain it though.

Best regards,

Akos

Has this bug been fixed in Marvin 5.3.3?

Hi Bob,

We have a long weekend due to our national holiday here, but Ákos will reply you soon.

Regards,
Erika.

Hi,

 

This fix is not included in the 5.3.3 version.

I fixed this bug. I think it will be available in the next patch release.

 

Jozsi