Thanks Greg, I can accept explanation for aromaticity but for structure 11 it will be good to have ability to read it with warning for example. All structure editors can read it!
Regards, Evgueni 2008/9/2 Greg Landrum <[email protected]> > Dear Evgueni, > > > > I have enclosed SDF file with some structures and file with problems. > > The main things is that SDMolSupplier cannot read structure #11 which any > > other tool will read without any problems and in some structures some > cycles > > ROMol will consider wrong as aromatic (see structures 3,4,5 for example) > - > > see Problems.doc for details. > > > > Any suggestions? > > I have not looked at all the structures with problems, but I don't see > anything here that's overly alarming. > > The aromaticity cases you refer to as "wrong" match the RDKit > definition of aromaticity : the ring c1nc(=O)cc[nH]1 is conjugated and > has 6 pi electrons (0 from the C with the extracyclic double bond to > O, 2 from the NH, 1 from the other atoms), so it's considered > aromatic. > > Structure 11 contains the substructure CN1=C[NH]N=C1, which contains a > 4-coordinate neutral nitrogen in a non-aromatic ring. This is not > considered a valid valence state for the N, so the RDKit won't process > the molecule. > > Best Regards, > -greg > -- Dr. Evgueni Kolossov (PhD) [email protected] Tel. +44(0)1628 627168 Mob. +44(0)7812070446
