Normally in MD simulations of biological molecules, one allows the geometry of molecules to be flexible in order to """"""""model"""""""" the system more realistically. But the atomic charges, which are used to describe intermolecular interactions, are fixed. My preliminary study indicates that flexible charges have an important impact on MD simulation results. I would like to carry this project further. For a large system, a technique has to be developed in which the charges are not chosen directly as a function of geometry, but as a function of time or some fragments of the molecule. For a small system, we can combine quantum chemistry and MD techniques together to do simulations. Now I am testing the MD/QM results on an isomerization reaction and a proton-transfer reaction.
Showing the most recent 10 out of 508 publications