On Tue, Oct 27, 2009 at 8:29 AM, Mark Rijnbeek <markr@ebi.ac.uk> wrote:
> Actually, I think this makes sense. The SQT does not know that the query
> is an actual molecule (i.e., completely defined SMILES). As a result,
> initialization of the query does not include aromaticity perception
> (which for SMARTS patterns in general, wouldn't be appropriate (?)).
> As a result, the query refers to an aliphatic C connected by a double
> bond to an aliphatic N - which is not the case in the target structure,
> after aromaticity has been perceived.
> Hence the failure to match

So if I wrap this in OrChem, the way to use the SQT to match on aromatic
compounds would be to instruct users to set aromaticity explicitly in
their query structure, like with c1ccncn1. If a user relies on
aromaticity detection (like you can with SMILES), things may not work.


Rajarshi Guha
NIH Chemical Genomics Center