Trouble with BondAngleChecker

Hi,

We are using JChem cartridge Java API 14.10.13.0.

If we process the attached mol file through the below Java code, it reports bad bond angles (no matter which constructor we use), even though the structure seems like it should pass ok (and we put this structure through Marvin 2D clean as well).  Is there a way to have the attached mol file not report bad bond angles?

Thanks for any help/advice.

Andrew

 

        MolImporter mi = new MolImporter(new File(in_mol), "mol");

        Molecule mol = mi.read();

        //BondAngleChecker bac = new BondAngleChecker(2.5);

        BondAngleChecker bac = new BondAngleChecker();

        StructureCheckerResult scr = bac.check(mol);

        if (scr!=null) {

            System.out.println("Bad bond angles found");

        }

        mi.close();

To add another piece of information .. the above Java and mol file do not cause reported bond-angle error in Jchem 5.11.2

Hi,

I'm moving the topic to the structurechecker forum, they will reply you as soon as possible.
Also it seems to me that cleaning in Marvin results in a slightly different structure. (attached). Can you confirm or refute this with your environment?

Regards, Domi

Thank you for changing forum.

And thank you for 2D cleaning result.  It is interesting:

   Marvin 2D clean in 5.12 create ba.mol that fail BondAngleChecker in API 14.10.13.0.

   Marvin 2D clean in 14.10.27.0 create result you attach (OH and Cl are closer together) and that pass BondAngleChecker in API 14.10.13.0.

   Both mol files pass BondAngleChecker in API 5.11.2, with either BondAngleChecker() or BondAngleChecker(2.5) constructor.

  In API 14.10.13.0 only 14.10.27.0-cleaned mol file pass BondAngleChecker.  Also, I cannot add any amount of tolerance (e.g. BondAngleChecker(250.0) ) to make it pass.

Can you provide suggestion?  After API update 5.12->14.10.13.0 users are suddenly getting many bond-angle errors.

The tolerance constructor in API 14.10.13.0 is ignored now, is that correct?

Thank you,

Andrew

 

Hi Andrew,

According to the
- tolerance question, no tolerance is taken into consideration as that functionality is deprecated and ignored
- angle errors, you can use the bond angle related fixer or the Clean2D itself to avoid bond angle errors
Is this OK for you?

Best regards,
Domi

Hi,

Thank you for the response.  In general, we prefer a more permissive structure check, so that we can leave structure as the chemist has drawn it without changes.

We just did not expect this new behavior, so I wanted to make sure it was intentional.

Thanks again and happy holidays!

Andrew