From: John M. <jo...@eb...> - 2013-11-11 15:42:28
|
Yes you need to process the structures separately. It’s not that there is a good but more likely there is non-covalent bonding - you can see how toolkits interpret - [Na+].[Cl-]. There might be some code in a down stream product which lays out multiple molecules and have CC’ed in cdk-devel list. J On 11 Nov 2013, at 15:26, Christophe Marcourt <chr...@ma...> wrote: > Ok, now it seems everything is working except that sometimes I have a CDKException due to disconnected molecule. So I'm using ConnectivityChecker.partitionIntoMolecules() method but I have a set of IAtomContainer. > Do I have to process all of the IAtomContainer or is there only one which represents the good structure ? And how to get the good one ? ;-) > Thank you. > > > Christophe Marcourt > > > On Mon, Nov 11, 2013 at 3:49 PM, Christophe Marcourt <chr...@ma...> wrote: > Ok, I already know this, so nothing more. > Fair enough. > Thank you for your time. > > > Christophe Marcourt > > > On Mon, Nov 11, 2013 at 3:48 PM, John May <jo...@eb...> wrote: > The JavaDoc has some… here’s how to create an IAtomContainer from an InChI string. > > http://cdk.github.io/cdk/1.5/docs/api/index.html?org/openscience/cdk/inchi/InChIToStructure.html > > and here shows generating coordinates > > http://cdk.github.io/cdk/1.5/docs/api/index.html?org/openscience/cdk/layout/StructureDiagramGenerator.html > > Cheers, > J > > On 11 Nov 2013, at 14:29, Christophe Marcourt <chr...@ma...> wrote: > >> Thank you very much. >> Is there a documentation which describes how to get a structure from an inchi using cdk java API ? >> >> >> Christophe Marcourt >> >> >> On Mon, Nov 11, 2013 at 3:25 PM, John May <jo...@eb...> wrote: >> Hi Christophe, >> >> The InChI stores coordinates in the AuxInfo and without that the IUPAC binary (called by the CDK) won’t set structure coordinates. To obtain a depiction you need to generate coordinates from scratch. The 2D coordinates are generated using the StructureDiagramGenerator whilst 3D coordinates are generated with ModelBuilder3D. It should be noted that neither method currently respects stereochemistry although that information is present. >> >> Thanks, >> John >> >> On 11 Nov 2013, at 12:59, Christophe Marcourt <chr...@ma...> wrote: >> >>> Hello all, >>> >>> I'm using cdk-inchi packages to turn an Inchi code to a strcuture. >>> I used the written code from the InChIToStructure javadoc. >>> It works fine but as the javadoc said "The generated IAtomContainer will have all 2D and 3D coordinates set to 0.0" ! >>> When I used chemspider web services, it returns with a good structure. >>> So how can one use InChIToStructure to have the real structure ? And not only with 0 in it... >>> Thanks. >>> >>> Christophe Marcourt >>> ------------------------------------------------------------------------------ >>> November Webinars for C, C++, Fortran Developers >>> Accelerate application performance with scalable programming models. Explore >>> techniques for threading, error checking, porting, and tuning. Get the most >>> from the latest Intel processors and coprocessors. See abstracts and register >>> http://pubads.g.doubleclick.net/gampad/clk?id=60136231&iu=/4140/ostg.clktrk_______________________________________________ >>> Cdk-user mailing list >>> Cdk...@li... >>> https://lists.sourceforge.net/lists/listinfo/cdk-user >> >> > > > |
From: Egon W. <ego...@gm...> - 2013-11-11 17:04:34
|
On Mon, Nov 11, 2013 at 4:42 PM, John May <jo...@eb...> wrote: > Yes you need to process the structures separately. It’s not that there is a > good but more likely there is non-covalent bonding - you can see how > toolkits interpret - [Na+].[Cl-]. I have not tried it, but it may be worth trying this: - add a IBond for salt bridges like this, e.g. with zero electrons, and a custom property marking it as a ionic bond - do the structure diagram generation - remove all bonds marked as ionic bond Egon -- Dr E.L. Willighagen Postdoctoral Researcher Department of Bioinformatics - BiGCaT Maastricht University (http://www.bigcat.unimaas.nl/) Homepage: http://egonw.github.com/ LinkedIn: http://se.linkedin.com/in/egonw Blog: http://chem-bla-ics.blogspot.com/ PubList: http://www.citeulike.org/user/egonw/tag/papers ORCID: 0000-0001-7542-0286 |
From: Christophe M. <chr...@ma...> - 2013-11-12 08:56:02
|
Thank you very much. The problem is that in my Inchi code I have something like 2C... for the molecular formule. A kind of symmetric molecule. So for the moment I'm requesting chemspider web service for this kind of molecule. Adding bonds and remove them later won't be a solution for me. Christophe Marcourt On Mon, Nov 11, 2013 at 6:04 PM, Egon Willighagen < ego...@gm...> wrote: > On Mon, Nov 11, 2013 at 4:42 PM, John May <jo...@eb...> wrote: > > Yes you need to process the structures separately. It’s not that there > is a > > good but more likely there is non-covalent bonding - you can see how > > toolkits interpret - [Na+].[Cl-]. > > I have not tried it, but it may be worth trying this: > > - add a IBond for salt bridges like this, e.g. with zero electrons, > and a custom property marking it as a ionic bond > - do the structure diagram generation > - remove all bonds marked as ionic bond > > Egon > > -- > Dr E.L. Willighagen > Postdoctoral Researcher > Department of Bioinformatics - BiGCaT > Maastricht University (http://www.bigcat.unimaas.nl/) > Homepage: http://egonw.github.com/ > LinkedIn: http://se.linkedin.com/in/egonw > Blog: http://chem-bla-ics.blogspot.com/ > PubList: http://www.citeulike.org/user/egonw/tag/papers > ORCID: 0000-0001-7542-0286 > |
From: Christophe M. <chr...@ma...> - 2013-11-11 15:44:45
|
Ok thank you very much. I have to check with my wife. She is the brain, I'm the hands ;-) Thanks again. Christophe Marcourt On Mon, Nov 11, 2013 at 4:42 PM, John May <jo...@eb...> wrote: > Yes you need to process the structures separately. It’s not that there is > a good but more likely there is non-covalent bonding - you can see how > toolkits interpret - [Na+].[Cl-]<http://apps.ideaconsult.net:8080/ambit2/depict?search=%5BNa+%5D.%5BCl-%5D> > . > > There might be some code in a down stream product which lays out multiple > molecules and have CC’ed in cdk-devel list. > > J > > > On 11 Nov 2013, at 15:26, Christophe Marcourt <chr...@ma...> > wrote: > > Ok, now it seems everything is working except that sometimes I have a > CDKException due to disconnected molecule. So I'm > using ConnectivityChecker.partitionIntoMolecules() method but I have a set > of IAtomContainer. > Do I have to process all of the IAtomContainer or is there only one which > represents the good structure ? And how to get the good one ? ;-) > Thank you. > > > Christophe Marcourt > > > On Mon, Nov 11, 2013 at 3:49 PM, Christophe Marcourt < > chr...@ma...> wrote: > >> Ok, I already know this, so nothing more. >> Fair enough. >> Thank you for your time. >> >> >> Christophe Marcourt >> >> >> On Mon, Nov 11, 2013 at 3:48 PM, John May <jo...@eb...> wrote: >> >>> The JavaDoc has some… here’s how to create an IAtomContainer from an >>> InChI string. >>> >>> >>> http://cdk.github.io/cdk/1.5/docs/api/index.html?org/openscience/cdk/inchi/InChIToStructure.html >>> >>> and here shows generating coordinates >>> >>> >>> http://cdk.github.io/cdk/1.5/docs/api/index.html?org/openscience/cdk/layout/StructureDiagramGenerator.html >>> >>> Cheers, >>> J >>> >>> On 11 Nov 2013, at 14:29, Christophe Marcourt <chr...@ma...> >>> wrote: >>> >>> Thank you very much. >>> Is there a documentation which describes how to get a structure from an >>> inchi using cdk java API ? >>> >>> >>> Christophe Marcourt >>> >>> >>> On Mon, Nov 11, 2013 at 3:25 PM, John May <jo...@eb...> wrote: >>> >>>> Hi Christophe, >>>> >>>> The InChI stores coordinates in the AuxInfo and without that the IUPAC >>>> binary (called by the CDK) won’t set structure coordinates. To obtain a >>>> depiction you need to generate coordinates from scratch. The 2D coordinates >>>> are generated using the StructureDiagramGenerator<http://cdk.github.io/cdk/1.5/docs/api/index.html?org/openscience/cdk/layout/StructureDiagramGenerator.html> whilst >>>> 3D coordinates are generated with ModelBuilder3D<http://cdk.github.io/cdk/1.5/docs/api/index.html?org/openscience/cdk/modeling/builder3d/ModelBuilder3D.html>. >>>> It should be noted that neither method currently respects stereochemistry >>>> although that information is present. >>>> >>>> Thanks, >>>> John >>>> >>>> On 11 Nov 2013, at 12:59, Christophe Marcourt <chr...@ma...> >>>> wrote: >>>> >>>> Hello all, >>>> >>>> I'm using cdk-inchi packages to turn an Inchi code to a strcuture. >>>> I used the written code from the InChIToStructure javadoc. >>>> It works fine but as the javadoc said "The generated IAtomContainer >>>> will have all 2D and 3D coordinates set to 0.0" ! >>>> When I used chemspider web services, it returns with a good structure. >>>> So how can one use InChIToStructure to have the real structure ? And >>>> not only with 0 in it... >>>> Thanks. >>>> >>>> Christophe Marcourt >>>> ------------------------------------------------------------------------------ >>>> November Webinars for C, C++, Fortran Developers >>>> Accelerate application performance with scalable programming models. >>>> Explore >>>> techniques for threading, error checking, porting, and tuning. Get the >>>> most >>>> from the latest Intel processors and coprocessors. See abstracts and >>>> register >>>> >>>> http://pubads.g.doubleclick.net/gampad/clk?id=60136231&iu=/4140/ostg.clktrk_______________________________________________ >>>> Cdk-user mailing list >>>> Cdk...@li... >>>> https://lists.sourceforge.net/lists/listinfo/cdk-user >>>> >>>> >>>> >>> >>> >> > > |