(731i) New Methods for Combining Experimental Data and Molecular Simulations into Hybrid Models

Creating models that are consistent with experimental data is an essential task in computational modeling. This is generally done by iteratively tuning the input parameters of a simulation to match experimental data. A new, alternative method is to bias a simulation to match experimental data leading to a hybrid model composed of the original model and biasing terms derived from experimental data. This second approach has been shown to be the unique minimal bias and I will describe methods to achieve this in molecular simulation. It can be done by modifying average values of collective variables or morphing the PMF of a simulation into a desired target. This new approach has been designed for cases where experimental data provides partial information but can be integrated with simulations for a more complete picture. The example systems considered in this talk are Lennard-Jones fluids, electrolyte solutions, and DFT simulations of water with solvated protons. The sources of experimental data for these systems comes from quantum chemistry calculations, bioinformatics data, and experiments.