E/Z and isotopes

User 870ab5b546

11-05-2005 21:07:24

ACE deals with isotopes in the following way: If a structure in a condition has an atom with a specific isotope, ACE requires that the response have that same isotope also for a match to occur; and, if a structure in a condition has an atom with no isotope specified, the response may specify an isotope there, and it will still match.





I have saved three structures into the database that differ only in their E/Z isomerism due to isotopic substitution (MOL files below). The JChem database search was able to distinguish these three structures, or I would not have been able to save the latter two after I saved the first one. However, when ACE compares any of these structures to the first one, it tells me they are identical. Is there a flag that we need to set to make sure that E/Z isomers due to isotopic substitution are treated as not equal?





Code:






  Marvin  05110517042D





  8  7  0  0  0  0            999 V2000


   -3.6561    2.1687    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -4.3706    1.7562    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -4.3706    0.2749    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -3.6561   -0.1376    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -2.9417    0.2749    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -2.9417    1.7562    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -2.3583    2.3395    0.0000 H   0  0  0  0  0  0  0  0  0  0  0  0


   -2.9604    1.0249    0.0000 H   0  0  0  0  0  0  0  0  0  0  0  0


  1  2  1  0  0  0  0


  1  6  2  0  0  0  0


  2  3  2  0  0  0  0


  3  4  1  0  0  0  0


  4  5  2  0  0  0  0


  6  7  1  0  0  0  0


  5  8  1  0  0  0  0


M  ISO  2   7   2   8   2


M  END











  Marvin  05110517052D





  8  7  0  0  0  0            999 V2000


   -3.6561    2.1687    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -4.3706    1.7562    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -4.3393    0.4625    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -3.6248    0.0500    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -2.9104    0.4625    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -2.9417    1.7562    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -2.1708    2.2145    0.0000 H   0  0  0  0  0  0  0  0  0  0  0  0


   -2.0542   -0.0376    0.0000 H   0  0  0  0  0  0  0  0  0  0  0  0


  1  2  1  0  0  0  0


  1  6  2  0  0  0  0


  2  3  2  0  0  0  0


  3  4  1  0  0  0  0


  4  5  2  0  0  0  0


  6  7  1  0  0  0  0


  5  8  1  0  0  0  0


M  ISO  2   7   2   8   2


M  END











  Marvin  05110517052D





  8  7  0  0  0  0            999 V2000


   -3.6561    2.1687    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -4.3706    1.7562    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -4.3706   -0.1000    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -3.6561   -0.5125    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -2.9417   -0.1000    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -2.9417    1.7562    0.0000 C   0  0  0  0  0  0  0  0  0  0  0  0


   -2.9208    0.9957    0.0000 H   0  0  0  0  0  0  0  0  0  0  0  0


   -2.9604    0.6187    0.0000 H   0  0  0  0  0  0  0  0  0  0  0  0


  1  2  1  0  0  0  0


  1  6  2  0  0  0  0


  2  3  2  0  0  0  0


  3  4  1  0  0  0  0


  4  5  2  0  0  0  0


  6  7  1  0  0  0  0


  5  8  1  0  0  0  0


M  ISO  2   7   2   8   2


M  END


ChemAxon a3d59b832c

12-05-2005 07:51:48

Hi Bob,





There is an option which handles E/Z matching. Please check out method setDoubleBondStereoMatchingMode(int):





http://www.jchem.com/doc/api/chemaxon/sss/search/Search.html#setDoubleBondStereoMatchingMode(int)





This method is available both in MolSearch and JChemSearch. PERFECT matching (duplicate checking at insert) is using DBS_ALL. When E/Z configuration checking is needed for a double bond depending on the above option, isomers are also considered.





Please check the Query Guide for some additional information:


http://www.jchem.com/doc/user/Query.html#stereobond





Best Regards,


Szabolcs

User 870ab5b546

12-05-2005 11:50:44

I believe we are already using the setDoubleBondStereoMatchingMode method to compare responses with reference structures. If we weren't, then trans-2-butene would match to cis-2-butene, correct?





ACE doesn't match trans-2-butene to cis-2-butene, but it does match trans-1,2-dideuteroethene to cis-1,2-dideuteroethene. (On the other hand, it does not match a response of ethene to either trans- or cis-1,2-dideuteroethene.)





The question is, when setDoubleBondStereoMatchingMode is set to DBS_ALL, and setExactIsotopeMatching is false, will that create the behavior that I am reporting? Because I think in our database search, setExactIsotopeMatching = true, and in comparisons, setExactIsotopeMatching = false. If this is the case, there is a bug. Even with setExactIsotopeMatching = false, JChem should be able to tell that a response of trans-1,2-dideuteroethene does not match a reference structure of cis-1,2-dideuteroethene, because the response lacks a D atom in the position specified by the reference structure.





I should note that we do not see this bug in 3D stereochemistry; that is, cis-1,2-dideuterocyclohexane and trans-1,2-dideuterocyclohexane are not matched to one another by ACE.

ChemAxon a3d59b832c

12-05-2005 16:46:36

Hi Bob,





You are right, it should not match. We will check that issue.





All the best,


Szabolcs

ChemAxon a3d59b832c

13-05-2005 10:50:58

Bob,





I fixed the bug, and both next JChem releases (minor: 3.0.12 and major 3.1) will contain the fix.





Thank you for reporting this.





All the best,


Szabolcs

User 870ab5b546

23-05-2005 15:32:51

Szabolcs wrote:
I fixed the bug, and both next JChem releases (minor: 3.0.12 and major 3.1) will contain the fix.


Thanks. When can we expect these releases (days, weeks, months, ...)?

ChemAxon 9c0afc9aaf

23-05-2005 16:37:12

Hi Bob,





JChem 3.0.12 is coming out this week, version 3.1 is expected in 1-2 months.





Best regards,





Szilard