Re: [Cdk-devel] errors are vanishing

[Rajarshi G. <rg...@in...>]

On Sat, 2007-04-14 at 11:20 -0400, Rajarshi Guha wrote:
> On Sat, 2007-04-14 at 17:36 +0530, sushil ronghe wrote:
> 
> 
> >                 Can you give an example SMILES and SMARTS query that
> >                 shows this problem?
> > 
> >  
> >          yes,for toluene the query [x2]-C returns the 7 hit.But the
> >         atoms comes out are only two.
> 
> Well if there are only 2 matching atoms, that is correct (see in the
> Daylight depictmatch web page).
> 
> The bigger problem is why are you getting 7 hits in the first place?
> Toluene, should only have 1 hit for the above SMARTS

I am still not entire clear as to the semantics of x<n>.

Does this mean:

An atom that is connected to n rings, so the methyl carbon in toluene
would match [x1]

OR

An atom that is part of n ring bonds, so the methyl carbon in toluene
does not match [x1]

It seems that currently, we are working according to the first
definition. But the first definition seems a little ambiguous for atoms
*in* a ring.
-------------------------------------------------------------------
Rajarshi Guha <rg...@in...>
GPG Fingerprint: 0CCA 8EE2 2EEB 25E2 AB04 06F7 1BB9 E634 9B87 56EE
-------------------------------------------------------------------
The only perfect science is hindsight.