Ulf Ryde
Grant: 600 000 SEK/ year, during 2 years.
Title
Development of methods to treat metal sites in ligand-binding calculations
Abstract
Calculating accurate estimates for the binding affinity of small molecules (e.g. drug candidates) to a macromolecular receptor is one of the most important challenges of computational chemistry. The most promising approach is free-energy perturbation (FEP), which normally gives a useful accuracy for pharmacy industry. However, for certain targets, the accuracy is much worse. In particular, metalloproteins in which the drug candidate binds to the metal provide a prominent problem for pharmaceutical applications. In this project, we will try to solve this problem with a hierarchy of computational approaches. First, we will develop and test two approaches to obtain a bonded molecular mechanics force field for metal sites based on quantum mechanical (QM) calculations. Second, we employ explicit combined QM and molecular mechanics (QM/MM) FEP simulations with different QM approaches. Third, we employ data from the latter simulations to develop a machine-learning force fields for metal sites that will be used for FEP calculations. Both the first and third approaches will involve extensive method and code development. We will apply the methods on metalloenzymes of central pharmaceutical interest, e.g. methionine aminopeptidase-2 and matrix metalloproteinase 12. The project is a tight collaboration between Lund University, Chalmers and AstraZeneca and will involve a PhD student.
Valera Veryazov
Title
Data-Inspired Engineering of Backstage Atoms in Quantum Chemistry
Abstract
Ab initio quantum chemistry is a key tool for modeling the electronic structure and properties of small molecules. However, for larger systems with practical importance, additional approximations are sometimes essential, as calculations may otherwise be infeasible. This project proposes using artificial, atom-like objects to model the less important parts of molecules or solids. The parameters of these objects are adjusted by minimizing the electron density differences in the more important regions of the system. The method will be developed and tested on a variety of systems in collaboration with several research groups.