clean 3D callenge

Hello





I am currently testing the clean3D function(s). I found some interesting cases where the fast and the fine builder fail. In the first case you can argue that c0 is not enough to handle such a special molecule. But in the second case c3 fails where c0 results in a reasonable 3D structure.





For now I have as workaround a builder chain, first c3 then c0 which results in reasonable structures for all cases. But I do not understand if I get c0 or c3 quality structures in the end...

Code:

public class Test {     public static void main(String[] args) throws Exception{         Molecule mol = new Molecule();                 mol = new MolHandler("C1C23CCCCC12CCCC3").getMolecule();         mol.clean(3, "c0[prehydrogenize]"); //c3 works fine here         System.out.print(mol.toFormat("sdf"));                 mol = new MolHandler("C1CC2CCC1CCC2").getMolecule();         mol.clean(3, "c3[prehydrogenize][timelimit]0[hyperfine]"); //c0 works fine here         System.out.print(mol.toFormat("sdf"));           //Workaround         mol.clean(3, "c3[prehydrogenize][timelimit]0[hyperfine]");         mol.clean(3, "c00");         //results in c3 or c0 quality strucutre?     } }

Dear Lorenz,





Thanks for your bug reports. We already noticed earlier a bug which resulted in corrupted "poked" ring systems. This is fixed in the development version which is expected to be part of the next release. In the current released version unfortunately either the c0 and the c3 clean can produce this error.





The hyperfine option currently can be very slow. We are also planning to make it faster in the next release.





The last Clean3D invocation however will do a complete c0 clean run to the given structure which will overwrite the results of the previous run.





When the fixes for the poked ring bug or the improvements in the hyperfine are released i will inform you in this topic.





Gabor