Efficient and Accurate Force Fields for Computer-Aided-Drug Design Computer-aided drug design is hampered by the limited accuracy in the force fields that describe the interactions between drug molecules, their targets, and shared environment. An automated protocol, adaptive force matching (AFM), is proposed for creating customized force fields that enable more efficient and accurate computational studies of the structure and function of drug candidates. Putting AFM in place will require the adaption of highly efficient techniques for sampling of configurations and quantum mechanical calculations of large systems. The approach will be developed and evaluated in a series of aims with increasing complexity. In the first two aims, force fields are generated for prediction of hydration free energies of small drug-like molecules and of the binding affinities for guest-host pairs. The performance of the method will be tested on examples from the recent SAMPL4 competition. In the last aim, force fields are created for simulations that allow one to identify nisin derivatives with improved solubility and stability at physiological pH as required for the design of nisin-based novel antibiotics. This class of drug is generally referred to as lantibiotics, and it carries the promise to address the increasingly severe problem of multi-drug resistant bacterial infection.
By developing a new method for creating atomistic models for guest-host interactions, the successful completion of this project will increase the accuracy of computational prediction of protein-ligand binding affinities; therefore increase the reliability of computer aided drug design. This in turn will reduce the cost for drug discovery and have a great impact on public health.
