No conformer and no dreiding energy from existing molecule

Hi, //edit 4x//


I have a molecule Bisbynin found in rice,


Bisbynin, a novel secondary metabolite from the fungus Stachybotrys bisbyi (Srinivasan) Barron


Tetrahedron letters; [0040-4039] De Silva yr:1995 vol:36 iss:12 pg:1997





and I created all stereoisomers with Marvin and saved them as SDF.





Published Bisbynin structure is


C[C@@H]1OCC2=C([C@H]1O)[C@@H](O)[C@@H]1O[C@]1(C\C=C(/C)C)[C@H]2O


or


InChI=1/C15H22O5/c1-7(2)4-5-15-13(18)9-6-19-8(3)11(16)10(9)12(17)14(15)20-15/h4,8,11-14,16-18H,5-6H2,1-3H3/t8-,11-,12+,13-,14-,15+/m0/s1


or


ICHJNTDKHBXTFN-CMZGOGIXBZ








However most of the calculations fail.


cxcalc with multiple options gives (includin 3d surface and conformer and dreiding all give an exception).


cxcalc dreidingenergy bisbynin-64-stereosiomers.sdf


cxcalc leconfonformer bisbynin-64-stereosiomers.sdf


Example:


leconformer: chemaxon.marvin.plugin.PluginException: No valid conformer or structure generation failed.





However minimum one compound should be available,


because the X-ray structure was published. Is there any solution


to that problem? This is Marvin 5.1.3.2





Attached all 64 Marvin generated stereoisomers from Bisbynin plus the picture itself. I also added the correct looking 3D structures from


smi23d just for comparison.


I also add the 64 unique InchIKeys generated with OpenBabel


with all hydrogens added.





ICHJNTDKHBXTFN-IJEOMIPRBW


ICHJNTDKHBXTFN-BLSUVFFWBK


ICHJNTDKHBXTFN-GIUROUPSBV


ICHJNTDKHBXTFN-SGFPCTINBV


ICHJNTDKHBXTFN-SEJAHTJSBZ


ICHJNTDKHBXTFN-SNSKAMRQBJ


ICHJNTDKHBXTFN-MBNVLLGTBZ


ICHJNTDKHBXTFN-BTTYOESUBU


ICHJNTDKHBXTFN-CBEPKRKSBW


ICHJNTDKHBXTFN-ASYYIYNTBF


ICHJNTDKHBXTFN-AMVLHOOLBF


ICHJNTDKHBXTFN-OQWJBGSFBW


ICHJNTDKHBXTFN-BUYINNJUBT


ICHJNTDKHBXTFN-KGOYKBJIBT


ICHJNTDKHBXTFN-BJQBPRFCBJ


ICHJNTDKHBXTFN-VIOGZJQCBY


ICHJNTDKHBXTFN-DHCPGLSIBN


ICHJNTDKHBXTFN-LSCFGXPWBT


ICHJNTDKHBXTFN-WAUGQOEABT


ICHJNTDKHBXTFN-YQRXSSPCBA


ICHJNTDKHBXTFN-BEQIJBKUBG


ICHJNTDKHBXTFN-ISUMBIBSBF


ICHJNTDKHBXTFN-XMGOOKLKBD


ICHJNTDKHBXTFN-QJJFXDRGBZ


ICHJNTDKHBXTFN-QCEUIWQLBU


ICHJNTDKHBXTFN-HOVKHMQVBX


ICHJNTDKHBXTFN-SEQNKZIQBF


ICHJNTDKHBXTFN-IPOPOUELBM


ICHJNTDKHBXTFN-HVBYWRBQBS


ICHJNTDKHBXTFN-RPADAGIIBO


ICHJNTDKHBXTFN-ORSZZKDPBC


ICHJNTDKHBXTFN-PWYDMPFYBX


ICHJNTDKHBXTFN-SBTCGRCQBW


ICHJNTDKHBXTFN-TVMMYHFSBX


ICHJNTDKHBXTFN-QVGMBTLQBP


ICHJNTDKHBXTFN-LNOQVRKOBV


ICHJNTDKHBXTFN-KDNBMIHNBE


ICHJNTDKHBXTFN-WIXLSBDIBL


ICHJNTDKHBXTFN-ULIKBCIWBB


ICHJNTDKHBXTFN-OSKCFZPOBC


ICHJNTDKHBXTFN-LSLMBRRXBH


ICHJNTDKHBXTFN-FANJIGQHBX


ICHJNTDKHBXTFN-SGEGXKEUBQ


ICHJNTDKHBXTFN-GCVRMMMFBM


ICHJNTDKHBXTFN-PGPFIFOTBA


ICHJNTDKHBXTFN-YUZSJEEMBL


ICHJNTDKHBXTFN-ZEISGHQPBW


ICHJNTDKHBXTFN-KFBSQEHXBN


ICHJNTDKHBXTFN-RWYPZYDQBV


ICHJNTDKHBXTFN-WYVUOJQUBZ


ICHJNTDKHBXTFN-JONGFPSMBV


ICHJNTDKHBXTFN-UJLMXNMRBE


ICHJNTDKHBXTFN-BOQOOYJYBX


ICHJNTDKHBXTFN-STSIKYPEBP


ICHJNTDKHBXTFN-CMZGOGIXBZ


ICHJNTDKHBXTFN-KQXWWFQMBP


ICHJNTDKHBXTFN-QDEFKWHGBG


ICHJNTDKHBXTFN-ZFHFLVAIBZ


ICHJNTDKHBXTFN-XJEGAWNGBG


ICHJNTDKHBXTFN-WDENSYTCBG


ICHJNTDKHBXTFN-DDGHMNHHBJ


ICHJNTDKHBXTFN-SIUZFRQLBL


ICHJNTDKHBXTFN-TXQXCMDOBJ


ICHJNTDKHBXTFN-WPRBHFCZBH








Sub-problem A)


Actually it would be nice if the molconverter could also generate


InChIKeys.





Sub-Problem (B)


Assume generating stereoisomers with marvin-mview is fine, however


saving those molecules directly to a file results in a random


number of those molecules. This is independent on the scrollbar


(either up or down to EOF). So in this case it would save 32, 48, 56


or 64 (correct) stereoisomers. That is really weird and the problem itself is


reproducible.








Cheers


Tobias (edited 3x)

Dear Tobias,





Thank You for the bug report. We will check it and answer in this topic.





Regards,


Gabor

Gabor,


in version 5.1.0 it worked! In 5.1.3.2 it doesn't work anymore.


Tobias

Dear Tobias,

 

Sorry for the late notification. It seems that valid 3d coordinates can be generated for all of the 64 stereosiomers (see attached file) now.

 

Your two further questions were missed - is it possible that you added them later?

A., you can use inchikey as output format in molconvert (see molconvert -Ho)

B,. i generated all 64 stereoisomers and "save all" worked fine. Could you clarify?

Sorry again for the late answer.

 

Regards

Gabor

Hi,

thanks alot. I can confirm that it works (I am still stuck in 5.2.2) but I will try the new 3D implementations as I have seen them in the SandBox and will update soon to the latest Marvin version. The other questions were added later but are resolved so far.

Cheers

Tobias