Yannick, The Aromaticity class is what you need. This assigns aromatic flags to the atoms and bonds of a structure.
> I need to kekulize it, which as you pointed leads to information loss. That's the wrong way round, the information loss is by storing an molecule without specified bond order (i.e. aromatic) in a molfile. The Kekulé from should be the default state of your structure and how it is stored on disk. You do not want or need to store aromaticity and its very cheap to assigned (i.e. with the above class). Cheers, J On Aug 12, 2014, at 7:11 PM, Yannick .Djoumbou <y.djoum...@gmail.com> wrote: > Hello, > > I have the same question as Zheng. The latest version of CDK returns exactly > the same results as the ones Zheng is getting. > An example where I see this might cause some problems is target prediction. > Some proteins/enzymes such as CYPs or UGTs can target groups where the > central atom is aromatic or is linked to an aromatic atom. What would be the > best way to deal with this kind of situation? > I notice that in order to study the structure of molecules in details, I need > to kekulize it, which as you pointed leads to information loss. > > Cheers, > > Yannick > > > On Fri, Aug 8, 2014 at 5:40 PM, John May <john...@ebi.ac.uk> wrote: > (on list) > > Hi Zheng, > > Aromaticity is stored as a flag on atoms and bonds (CDKConstants.ISAROMATIC). > When given an aromatic compound one needs to kekulise or dearomatise the > structure. There are multiple ways of doing this (see the API) the most > recent is the ‘Kekulise’ class. > > However one problem here is you have information loss. The aromatic bonds are > a query feature of the ctfile and are not good to represent non-query > structures. It’s a bit long to explain but basically aromaticity in a mol > files is ambiguous. Note that SMILES and InChI automatically have > single/double bond orders assigned on input because they are unambiguous*. > ChemAxon have added an extension to the molfile allow this (see here) but > that is specific to them and to my knowledge not widely used. > > As such, the molfile reader(s) will set the bond order to single and the > aromatic flag. Out of interest, which mdl reader are you using? The > MDLV2000Reader should actually be setting bond order = Unset? > > Some links that would be of use > > http://tech.knime.org/forum/cdk/aromaticity-in-cdk > http://chem-bla-ics.blogspot.com.es/2011/10/cdk-file-formats-1-mdl-molfiles-and.html > http://blueobelisk.shapado.com/questions/detect-aromatic-rings-using-bond-order-4-in-a-sdf-file-with-cdk > <— shows example of getting bond orders from a molfile > > Hope that helps, > John > > *well not always true but that throws an error > > On Aug 8, 2014, at 9:28 PM, Zheng Shi <zs...@ualberta.ca> wrote: > >> Hi, >> >> I just wonder how I can get the bond type for benzene in cdk. I can't find >> function to get bond type, but have function to get the bond order, so I >> just wonder whether bond order is equivalent to bond type. And for the bond >> order types, it seems have four types:single, double. triple and quadruple. >> Then how about the aromatic bond in benzene? It seems you will get different >> results for different representations for benzene. If you represent benzene >> as single, double bonds as the picture, it will return single ,double for >> the bond order. But if you represent as aromatic form, it will only return >> single as the bond order, not aromatic bond. I don't know why. I just wonder >> how to get bond type, is it equivalent to bond order? And what about the >> aromatic bond? Why does cdk return single, not aromatic bond? How to get a >> aromatic bond type? Thanks. >> <Capture1.JPG><Capture2.JPG>------------------------------------------------------------------------------ >> Want fast and easy access to all the code in your enterprise? Index and >> search up to 200,000 lines of code with a free copy of Black Duck >> Code Sight - the same software that powers the world's largest code >> search on Ohloh, the Black Duck Open Hub! Try it now. >> http://p.sf.net/sfu/bds_______________________________________________ >> Cdk-user mailing list >> Cdk-user@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/cdk-user > > > ------------------------------------------------------------------------------ > > _______________________________________________ > Cdk-user mailing list > Cdk-user@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/cdk-user > >
------------------------------------------------------------------------------
_______________________________________________ Cdk-user mailing list Cdk-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/cdk-user
