Structure Search Question

Using the latest version jchem 5.1.04, I wonder why the attached Query (MRV file) does not match target:


CN1C2=C(C=CC=C2)N3N=C(C)C(C(C)=O)=C13




Thanks.

Target: CC(C)C1=CC(C)=C(Cl)C=C1




Query: see attachment. I specify the bound as single or double.




There is no match using jchem_5_1_04

Hi,





Unfortunately both of your MRV files are 0 bytes.


Could you try again please (you can also edit your post) ?


If it seems to be a forum bug we will fix it ASAP ... please paste the mrv sources until then as text.





Best regards,





Szilard

First query








<?xml version="1.0" ?>


<cml>


<MDocument>


  <MChemicalStruct>


    <molecule molID="m1">


      <atomArray


          atomID="a1 a2 a3 a4 a5 a6 a7 a8 a9 a10 a11 a12 a13 a14 a15 a16 a17"


          elementType="C C O C C C N N C C C C C C N C C"


          x2="14.218960000000003 15.552693333333334 15.552693333333334 16.88642666666667 18.29314666666667 18.61346666666667 19.323733333333337 18.553733333333337 18.714640000000003 19.960453333333334 19.799546666666668 18.39264 17.14682666666667 17.307733333333335 16.27733333333334 14.745733333333336 17.047333333333338"


          y2="-4.7824 -4.012400000000001 -2.4724000000000004 -4.7824 -4.155946666666667 -2.6497333333333337 -5.300400000000001 -6.634133333333334 -8.165733333333336 -9.07088 -10.602480000000002 -11.228933333333336 -10.3236 -8.792186666666668 -7.647733333333335 -7.808640000000001 -6.314000000000001"


          />


      <bondArray>


        <bond atomRefs2="a1 a2" order="1" />


        <bond atomRefs2="a2 a3" order="2" />


        <bond atomRefs2="a2 a4" order="1" />


        <bond atomRefs2="a4 a5" order="1" />


        <bond atomRefs2="a5 a6" order="1" />


        <bond atomRefs2="a5 a7" order="2" />


        <bond atomRefs2="a7 a8" order="1" />


        <bond atomRefs2="a8 a9" order="1" />


        <bond atomRefs2="a9 a10" order="1" queryType="Any" />


        <bond atomRefs2="a10 a11" order="1" queryType="Any" />


        <bond atomRefs2="a11 a12" order="1" queryType="Any" />


        <bond atomRefs2="a12 a13" order="1" queryType="Any" />


        <bond atomRefs2="a13 a14" order="1" queryType="Any" />


        <bond atomRefs2="a9 a14" order="1" queryType="Any" />


        <bond atomRefs2="a14 a15" order="1" />


        <bond atomRefs2="a15 a16" order="1" />


        <bond atomRefs2="a15 a17" order="1" />


        <bond atomRefs2="a4 a17" order="2" />


        <bond atomRefs2="a8 a17" order="1" />


      </bondArray>


    </molecule>


  </MChemicalStruct>


</MDocument>


</cml>

Second Query





<?xml version="1.0" ?>


<cml>


<MDocument>


  <MChemicalStruct>


    <molecule molID="m1">


      <atomArray


          atomID="a1 a2 a3 a4 a5 a6 a7"


          elementType="C C C C C C C"


          x2="17.764999389648438 16.43133004860018 16.43133004860018 17.764999389648438 19.098668730696694 19.098668730696694 17.764999389648438"


          y2="-8.1399664240577 -8.90998336461674 -10.450017245734822 -11.220034186293862 -10.450017245734822 -8.90998336461674 -6.5999664240577"


          />


      <bondArray>


        <bond atomRefs2="a1 a2" order="1" />


        <bond atomRefs2="a2 a3" order="2" />


        <bond atomRefs2="a3 a4" order="1" />


        <bond atomRefs2="a4 a5" order="2" />


        <bond atomRefs2="a5 a6" order="1" />


        <bond atomRefs2="a1 a6" order="2" />


        <bond atomRefs2="a1 a7" order="1" queryType="SD" />


      </bondArray>


    </molecule>


  </MChemicalStruct>


</MDocument>


</cml>

Please let me try to give you a quick answer, one of my colleagues may give you a more detailed answer tomorrow.





The two cases seem to be similar.





In the first case you have replaced the bonds of a ring in fused aromatic system. This means that one of the bonds of the adjacent aromatic ring has also changed to any bond, while it's other bonds remained the same.


Therefore the bonds of this adjacent ring are not converted to aromatic during standardization (you can try than in Marvin Sketch).


If you replace the 3 non-converted bonds with "Single or Aromatic" the wury should find targets with either aromatic rings or single bonds in the middle ring.





In the case of the second query a single-or-double bond is attached to a benzene ring, and it seems to prevents aromatization - I guess interferes with Huckel's rule or some other rule applied.





Please see:


http://www.chemaxon.com/jchem/doc/user/query_standard.html#aromatization





and





http://www.chemaxon.com/jchem/doc/user/query_searchoptions.html#vaguebond





We will get back to you with more details and/or corrections.





Best regards,





Szilard

Hi,





First molecule (any bond case):








Yes, the reason is basically that the ANY bond prevents aromatization of the middle ring. In 5.2, we will be introducing a new query aromatization method, that means that such cases will be found.





Second molecule (single or double bond case):





I see two problems here:





1. Single or double bonds don't match aromatic bonds by definition. You have to use another query bond type (e.g. ANY or "single or aromatic" or "double or aromatic").





Alternatively, you could use one of the higher level vague bond settings to match aromatic bonds there also. (See vague bond levels 3 or 4.)


http://www.chemaxon.com/jchem/doc/user/query_searchoptions.html#vaguebond





2. As Szilárd mentioned, indeed the phenyl ring is not aromatized because of the "single or double bond". I guess this is because in the "double" case it would mean an invalid valency scenario which is not treated as aromatic. (The same happens when the "single or double" bond is replaced by an "any" bond.) However, I think that in case of these query bonds we still should treat the benzene ring aromatic. We will look into this question and try to fix it as part of query aromatization. The vague bond option is a valid workaround for this query also.





Best regards,





Szabolcs

Thanks for the explanations, all make sense, although the behavior is a little bit unintuitive from the end user point of view.  I will ask chemists try to explicitly aromatize their query structure if possible, when vague bonds are used.

Hi,





As I mentioned, improvements are on the way. We will also think about improving the single/double case...








Best regards,





Szabolcs

All the above cases find now their corresponding targets in the 5.3.X versions.

 

Best regards,

Szabolcs