search

TinkerTools / tinker9

Tinker9: Next Generation of Tinker with GPU Support
Other
48 stars 27 forks source link

Solvate command #202

Closed eneas77 closed 2 years ago

eneas77 commented 2 years ago

Hi,

I´m trying to use the Solvate command to make a gb minimization of my molecule, however, it looks like the enrgy is the same as the gas-phase results... Maybe is not clear for me the proper usage of the solvate KEYWORD, so I would appreciate a quick refresh of it at.

I just add the following line to the .key file: solvate gb

Is there anything more to add here?

Thank you kindly,

E77

jayponder commented 2 years ago

Generalized Born (keyword "solvate gb") is only implemented for fixed atomic partial charge force fields. In current Tinker, this would be the Amber, CHARMM, OPLS and MMFF variants. It is not available for MM2, MM3 or AMOEBA, and you will just get the "gas phase" energy if you try to use GB with these force fields. For AMOEBA there is an analogous implicit solvation method for polarizable atomic multipoles, Generalized Kirkwood, that is available via the keyword "solvate gk". Note that Tinker implements several small variations on these GB and GK methods, available via other keywords and as detailed in the user's guide. But I suggest you just use "solvate gb" or "solvate gk" unless you understand the specific small differences between methods.

jayponder commented 2 years ago

Sorry, I just saw that this was in the Tinker9 (ie, GPU code) discussion. I believe that implicit solvation models are not currently implemented at all in the Tinker9 GPU code. My comments directly above about use of GB and GK models apply only to the CPU version of Tinker (ie, "Tinker8").