Hi, I've taken a longer look at the problem, and I think that it is indeed the fault of the signature code. For the two versions of the test molecule we get:
...[C]([C]p[C]p[C]), [C](p[C]p[C]), [C](p[C]p[C]), [C](p[C]p[C]), [C](p[C]p[C]), [C](p[C]p[C][Cl])... for "ErnestBupropion" and : ...[C](p[C]p[C]), [C](p[C]p[C]), [C](p[C]p[C]), [C](p[C]p[C]), [C](p[C]p[C][C]), [C](p[C]p[C][Cl]),... for "MeteorBupropion". These are just the sublist of signatures for the atoms that have aromatic bonds (at height 2). The error here is for "[C]([C]p[C]p[C])" and "[C](p[C]p[C][C])" which should not be different. The 'p' stands for aromatic, if you didn't already guess :) So this might take a little while to fix, assuming there is no quick patch. Perhaps next weekend or so - then I will upload a new version of the library as a patch to the CDK, and then announce to this list. Probably discussion of the problem (if any) should go on in cdk-devel list, although people are welcome to mail me directly, or even fix it themselves if they want to! gilleain On 3/15/13, Patrik Rydberg <[email protected]> wrote: > Hi John, > > Thanks for the tip, but I'm already using the toCanonicalString function, so > this should not be happening. Which is somewhat troubling. > > Gilleain Torrance is looking at the code as well right now. > > Patrik > > -------------------------------------------------------------------------- > Patrik Rydberg, Associate Professor, Ph.D. > > Biostructural Research Group > Department of Drug Design and Pharmacology > Faculty of Health and Medical Sciences > University of Copenhagen > DK-2100 Copenhagen > > http://www.farma.ku.dk/index.php/Patrik-Rydberg/5304/0/ > > From: John May [mailto:[email protected]] > Sent: 14. marts 2013 18:30 > To: Patrik Rydberg > Cc: [email protected] > Subject: Re: [Cdk-user] problems in the use of the signature class > > Hi Patrik, > > This might help, > https://github.com/egonw/cdk/blob/master/src/main/org/openscience/cdk/fingerprint/SignatureFingerprinter.java#L73-L80 > > I can have a look for if you like, have you got the code? > > J > > On 14 Mar 2013, at 12:49, Patrik Rydberg > <[email protected]<mailto:[email protected]>> wrote: > > > Hi CDK users, > > Is there anyone who has a working code that creates atomic signatures of > varying height for a molecule? > > I have made my own code, but for some strange reason it can on occasion > generate slightly different signatures for the same molecular structure > (same molecule, but two different mol files with different atom ordering). > I'm quite confused in how this can happen and wondering if the error is in > my code or in the signature class. > > If anyone would be willing to help me find the error I would be very > grateful. > > Best regards, > Patrik Rydberg > > > -------------------------------------------------------------------------- > Patrik Rydberg, Associate Professor, Ph.D. > > Biostructural Research Group > Department of Drug Design and Pharmacology > Faculty of Health and Medical Sciences > University of Copenhagen > DK-2100 Copenhagen > > http://www.farma.ku.dk/index.php/Patrik-Rydberg/5304/0/ > > > ------------------------------------------------------------------------------ > Everyone hates slow websites. So do we. > Make your web apps faster with AppDynamics > Download AppDynamics Lite for free today: > http://p.sf.net/sfu/appdyn_d2d_mar_______________________________________________ > Cdk-user mailing list > [email protected]<mailto:[email protected]> > https://lists.sourceforge.net/lists/listinfo/cdk-user > > ------------------------------------------------------------------------------ Everyone hates slow websites. So do we. Make your web apps faster with AppDynamics Download AppDynamics Lite for free today: http://p.sf.net/sfu/appdyn_d2d_mar _______________________________________________ Cdk-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/cdk-user
