Re: [Rdkit-discuss] Strange MMFF94 Optimization Results?

I started using the SD format instead, and it fixed my problems (the molecules now have correct bond orders and they are MMFF94 optimized properly). Thanks for the help! On Thu, Nov 2, 2017 at 3:21 PM, Paolo Tosco wrote: > Hi Patrick, > > I don't know Avogadro in detail,

Re: [Rdkit-discuss] Strange MMFF94 Optimization Results?

Hi Patrick, I don't know Avogadro in detail, but as it is based on OpenBabel I can imagine it uses an algorithm to guess bond orders from bond distances, angles, etc., whereas the RDKit does not. In fact, if I convert a benzene ring from SMILES to SDF through PDB format (which does not

Re: [Rdkit-discuss] Strange MMFF94 Optimization Results?

Yes, that is probably correct. I am loading a pdb file for the initial conformer, so bond orders are not specified. But I find it strange, still, that when I use MMFF94 in Avogadro to optimize it, it results in a planar shape even though all the bonds are still single. On Thu, Nov 2, 2017 at

Re: [Rdkit-discuss] Strange MMFF94 Optimization Results?

Dear Patrick, my guess is that you loaded the caffeine coordinates from PDB, or anyway from a format where bond orders were not specified. All atoms appear to be sp3-hybridized, which results in the wrong geometry being generated. Hope that helps, Paolo On 11/02/17 18:20, Patrick Avery