Hi,

On Sat, Feb 20, 2010 at 4:22 PM, Noel O'Boyle <baoille...@gmail.com> wrote:
> On 20 February 2010 13:35, Geoffrey Hutchison <ge...@geoffhutchison.net> 
> wrote:
>>> http://my.cdash.org/index.php?project=Open%20Babel
>>
>> Here's another failure. This is on the SMILES/SMARTS match test (i.e., all 
>> of these SMILES should match themselves when turned into SMARTS).
>>
>> O([P@@](N)(=O)c1ccccc1)CC
>>
>> When I read this in as a SMILES and output as a SMILES, I lose chirality:
>>
>> O(P(=O)(N)c1ccccc1)CC
>>
>> Now if I look at the smartsmatch results 
>> (http://my.cdash.org/testDetails.php?test=1816390&build=49366) I can see 
>> that each of the failing molecules has a "chiral" non-carbon atom like 
>> "c...@]1[c@@H](CCC1)[C@@H]1CCCNC1".
>>
>> In previous versions, we retained these chiral specifiers. I can certainly 
>> understand "N" being ignored, but "P" and "S" can definitely be chiral. Does 
>> the new stereo analysis ignore all non-carbon atoms for chirality?
>
> I think the answer is probably "yes". I'll look into it. I have a
> feeling that there is some corner case where a lone pair is involved
> in the chirality, but it should be relatively to deal with the general
> case.

Yes, N, P and S are currently skipped. Changing the line below (259,
src/stereo/perception.cpp) makes the smilesmatch test pass (see
Experimental Dashboard). However, I'm not 100% sure what would be the
best option for nitrogen. Detecting chiral (bridge-head) nitrogen
atoms seems complex...

      // skip N, P, S
      if (atom->IsNitrogen() || atom->IsPhosphorus() || atom->IsSulfur())
        continue;

Tim

>> Thanks,
>> -Geoff
>>
>> (Two tests with clear analysis, two tests "failing" due to bug-fixes which 
>> have been updated, one left)
>> ------------------------------------------------------------------------------
>> Download Intel&#174; Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev
>> _______________________________________________
>> OpenBabel-Devel mailing list
>> OpenBabel-Devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/openbabel-devel
>>
>

------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
OpenBabel-Devel mailing list
OpenBabel-Devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/openbabel-devel

Reply via email to