Re: [Cdk-devel] Wow!

[Pantelis S. <ch...@ma...>]

Hi Rajarshi,
  Yes, true. I was not expecting fingerprints to be unique, but I was
really amazed at the ease with which I found such "counterexamples". I
tried extended fingerprints over a set of around 2000 molecules and
there were no identical fingerprints. Would be very interesting to find
compounds that have all, let's say, CDK-provided descriptors equal but
are different (I'm almost sure there must me some already in some
database out there).

Regards,
Pantelis

On Tue, 2012-05-15 at 09:41 -0400, Rajarshi Guha wrote:
> It also depends on what fingerprint your using - the extended
> fingerprints will differentiate between them at the default settings.
> 
> In the end appropriate choice of representation (fingerprint or
> otherwise) can make any two molecules look identical or different :)
> 
> On Tue, May 15, 2012 at 8:35 AM, Pantelis Sopasakis <ch...@ma...> wrote:
> > Hello everyone,
> > Something not really relevant to CDK... Look what I found playing with
> > fingerprint calculation. The following compounds:
> >
> > * naphthalene
> > * phenanthrene
> > * hexacyclo[12.12.0.0^{2,11}.0^{5,10}.0^{16,25}.0^{18,23}]
> > hexacosa-1(26),2,4,6,8,10,12,14,16,18,20,22,24-tridecaene
> >
> > (see attached SDF files)
> >
> > have identical fingerprints:
> >
> > {62, 95, 161, 213, 253, 463, 535, 574, 742}
> >
> > Regards,
> > Pantelis Sopasakis, ch...@ma...
> > Dipl. Chemical Engineer NTUA
> > Msc. Applied Mathematics NTUA
> > Automatic Control Unit, School of Chem. Eng., NTUA
> > Heroon Polytechneiou 9, 15780 Zografou Campus, Athens, Greece
> > Tel. (+30) 210 772 3236 (office)
> >
> > ------------------------------------------------------------------------------
> > Live Security Virtual Conference
> > Exclusive live event will cover all the ways today's security and
> > threat landscape has changed and how IT managers can respond. Discussions
> > will include endpoint security, mobile security and the latest in malware
> > threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> > _______________________________________________
> > Cdk-devel mailing list
> > Cdk...@li...
> > https://lists.sourceforge.net/lists/listinfo/cdk-devel
> >
> 
> 
>