The ability to perform large-scale first principles quantum mechanical calculations on entire biomolecules such as proteins and DNA provides a unique tool for accurate microscopic insights on the functions and properties of these biomolecules.
Our work in this area focuses on using quantum calculations to accurately predict the interactions between biomolecules and other molecules (ligands) with applications in drug design. Using Energy Decomposition Analysis (EDA) methods which we develop (J. Chem. Theory Comput. 12 (2016) 3135-3148; doi: 2010.1021/acs.jctc.6b00272 ), we can dissect the interaction between a biomolecule and a potential drug molecule into chemically relevant contributions. We can use this information to suggest chemical modifications to the ligand that can eventually be used to improve its action as a drug. For example, in the Thrombin protein (J. Chem. Theory Comput. 13 (2017) 1837-1850; doi: 10.1021/acs.jctc.6b01230?journalCode=jctcce ), which is a target for treatment of thrombosis disease, our preliminary EDA calculations allow us to rationalise and tailor the interactions of a variety of ligands.
Figure: A Thrombin-ligand complex (about 5000 atoms) on which EDA calculations have been performed with ONETEP. The active site of the protein with the bound ligand are depicted in the magnification.
Also very important for drug design is to be able to predict the free energies of binding of drugs to proteins and other biomolecular targets. Large-scale quantum calculations can lead to improvements in the accuracy of such predictions. However, free energy calculations are extremely challenging as we need to sample the dynamical behaviour of the protein to capture the entropic contributions to binding, and include the effects of the solvent and temperature. To overcome these obstacles we are developing multiscale methods by combining large-scale first principles quantum mechanical calculations with classical force fields and solvent models (Proteins 82 (2014) 3335-3346; doi: 10.1002/prot.24686 ; J. Phys. Chem. B 119 (2015) 7030-7040; doi: 10.1021/acs.jpcb.5b01625 ). This allows us to study the dynamical behaviour of biomolecules as well as the electronic polarization and charge transfer that determine all the interactions, including chemical bond formation. The ability to accurately describe these effects is crucial for the rational development of new drugs or the understanding of catalysis in enzymes.
Figure: Structure of the T4 Lysozyme protein (about 2500 atoms) from a quantum mechanical calculation in the presence of solvent (water). The solvation cavity surrounding the protein is also shown.