search

TinkerTools / poltype2

Poltype 2: Automated Parameterization and Free Energy Prediction for AMOEBA
Other
41 stars 11 forks source link

Suspicious asignment of polarizabilities #15

Closed vojtechkostal closed 2 years ago

vojtechkostal commented 2 years ago

Hello,

I recently used your Poltype2 code to parametrize an organic molecule, namely N-methylsulfamate. However, when I check the final.key file containing the parameters, I see comments that poltype2 identified the three terminal oxygens with sulfur atoms (sulfur and the terminal oxygens indeed share the same polarizability parameter). The same situation happened also to the nitrogen atom that was assigned with parameters from hydrogen on non-aromatic nitrogen and the carbon atom that was identified as hydrogen on non-aromatic carbon.

Could you please help me to figure out where could be the potential issue? As I got it correctly, polarizabilities are taken from your database and only multipoles undergo the fitting to the qm-based electrostatic potential (in my parametrization I can see unique set of multipoles at each atom).

I also attach an archive containing the initial geometry in .sdf format, .ini file and the final.key file the outputted parameters. poltype2-NMSM.zip

Thank you in advance for any advice.

misterbrandonwalker commented 2 years ago

Thank you for being vigilant and checking results to make sure they make sense. This is a somewhat more serious issue/bug introduced a few months ago while I was trying to handle a special case of polarizability assignment for molecules that are charged and also have resonance. It works fine now.