Re: [Rdkit-discuss] Valance Error on Aromatic N atoms.

2017-07-12 Thread Greg Landrum
The molecule has been constructed without problems, but the jupyter rendering code is failing. This isn't particularly surprising, that code is really set up to render non-query molecules, not things constructed from SMARTS. If you want to get a rendering of a molecule build from SMARTS in the

Re: [Rdkit-discuss] Valance Error on Aromatic N atoms.

2017-07-11 Thread Chicago Ji
Hi, Greg. Thanks a lot for your code. I met another error. I don't know whether it's Normal. I have a correct SMILES, and read it through MolFromSmarts: sa = 'CC(C)(C)c1ccc(cc1)C(O)CCCN1CCC(CC1)C(O)(c1c1)c1c1'

Re: [Rdkit-discuss] Valance Error on Aromatic N atoms.

2017-07-11 Thread Greg Landrum
Here's a simple bit of code that shows how to get only pharmacophore features whose atoms are contained completely in the MCS: https://gist.github.com/greglandrum/faf47bd161d9b3af5374c358d0a0786d An interesting extension to this, which I may do for a blog post later, is to further filter things

Re: [Rdkit-discuss] Valance Error on Aromatic N atoms.

2017-07-11 Thread Greg Landrum
Interesting use case. I can send some sample code for this later this afternoon. On Tue, Jul 11, 2017 at 11:43 AM, Chicago Ji wrote: > Hi, Greg, > > The SMARTS comes from MCS.FindMCS() analysis. It seems that MCS can only > output the result as SMARTS. > > Maybe I can

Re: [Rdkit-discuss] Valance Error on Aromatic N atoms.

2017-07-11 Thread Chicago Ji
Hi, Greg, The SMARTS comes from MCS.FindMCS() analysis. It seems that MCS can only output the result as SMARTS. Maybe I can calculate atom features of the "mother molecule', and then extract those belonging to the common substructure. Or I can delete some atoms of the 'mother molecule', and

Re: [Rdkit-discuss] Valance Error on Aromatic N atoms.

2017-07-11 Thread Greg Landrum
Using the pharmacophore search code on molecules constructed from SMARTS is, in many circumstances, unlikely to yield the results that you are looking for. Query-Query matching is different than Query-Atom matching. It's more common to look for ph4 features in molecules constructed from SMILES or

Re: [Rdkit-discuss] Valance Error on Aromatic N atoms.

2017-07-11 Thread Chicago Ji
Hi Greg, I want to get pharmacophore feature for each atom on the molecule using GetFeaturesForMol(). It seems that sanitization is a must before running GetFeaturesForMol(). Many Thanks. Best, Changge On Tue, Jul 11, 2017 at 7:31 AM, Greg Landrum wrote: > Hi

Re: [Rdkit-discuss] Valance Error on Aromatic N atoms.

2017-07-11 Thread Greg Landrum
Hi Changge, Is there any particular reason that you're calling updatePropertyCache on the molecule? That's not an operation that's normally necessary. Best, -greg On Mon, Jul 10, 2017 at 8:37 PM, Chicago Ji wrote: > Dear all, > > I met a problem with Aromatic N atoms.