Yang Mingjun (TALK)
More accurate molecular mechanics force field development is always one of the central problems in computer simulation of complicated heterogeneous systems. Although implicit solvation effect is included in the parametrization of various non-polarizable force fields, many properties cannot be well reproduced computationally. Based on the Drude model, the polarizable force field has been developed for carbohydrate systems. Furthermore, we also proposed new enhanced sampling method to study the conformation heterogeneity of complicated glycoproteins, the new findings from which harness our understanding on vaccine design targeting Salmonella and HIV.
Besides parametrization for biomolecules, we are also working on the development of a new general force field aiming to accurately cover at least one million drug-like chemical scaffolds. In contrast to the philosophy of general force field development, we are developing system specific force field to maximize the accuracy of one given system and give less weight for parameter transferrability to other chemical similar molecules. This force field has been successfully used in our prediction of crystal structure and solid state property for various forms, e.g. anhydrous, hydrate, salt and cocrystal forms.