Dear Sarath,

You correctly said that amber99sb forcefield does not account for the polarizability factor. However, in some cases the said factor plays crucial role to get the true free energy surface of a binding process. There are two ways available to tackle this problem.

One is to go with polarizable forcefield for protein as well as for water (Such as martini). However, it is quite expensive for bigger systems.

In other way, you can use mean field polarizable TIP3P model for water (commonly known as MFP/TIP3P model) with amber03 forcefield. It incorporates the polarizability of water on the fly within the non-polarizable framework of gromacs v.4.0.7. (please refer to https://arxiv.org/ftp/arxiv/papers/1505/1505.01454.pdf). It scales the atomic charges of water molecule (dipole of water) with respect to the changes of local environment. This correction does not add any extra computational burden into the calculations. We have used the aforementioned methodology in our recently published article ("Divalent ions are potential permeating blockers of the non-selective NaK ion channel: combined QM and MD based investigations" DOI: 10.1039/C7CP05586B). Hope it helps.

Regards,

Biswajit

Dear Biswajit Sadhu ,

I am glad to see your suggestion to use MFP/TIP3P water model with amber03 forcefield.

Recently, I have been searching for MFP/TIP3P forcefield file and the protocol to use it in GROMACS. Unfortunately, I am unable to locate them in the GROMACS manual and related documents.

I was wondering if you could help me by sharing a link or document or forcefield files.

I am hoping to get a reply.

Regards,

Sudarshan