Skip to main content
KBS_Icon_questionmark link-ico


The formation and cleavage of phosphate bonds are essential in most biological processes including reproduction, protein synthesis, signal transduction, energy storage, and transfer. We use computational methods to help reveal the molecular mechanisms of how enzymes can achieve both a high efficiency and specificity catalysing phosphate hydrolysis and transfer reactions.

We developed quantum-classical QM/MM simulation techniques combining Hamiltonian replica exchange with the finite temperature string method. These methods accurately model bond breaking and forming using ab initio calculations and also provide an enhanced sampling of protein conformations. We develop further novel computational methods that are applicable to accurately estimate enzyme activities in biomolecular complexes.

Metal ions are required for many enzymes and ribozymes to function. One- and two-metal ion catalysis are ubiquitous mechanisms employing metal ions in the centre of the active sites. Examples include kinases and phosphatases involved in signalling, polymerases and RNases involved in gene expression, and ATPases involved in molecular machines. We study the roles of metal ions in the catalytic mechanism, in particular, the changes in metal ion coordination and their roles in proton transfer processes coupled to the chemical step. Our research is strongly motivated by the potential applications to designing new drugs for diseases such as cancer and AIDS, by exploiting the possibility of direct binding between inhibitor and metal ion.

In addition to atomistic molecular simulations, high-level models of interacting biomolecules offer a complementary, top-down approach to integrate detailed atomistic information with network models of activities and interaction patterns of biomolecular complexes. We aim to use kinetic network models to characterise the complex protein-protein interactions that drive signaling pathways by determining kinase activities. Recently we applied kinetic network models to analyse the efficiency of enhanced sampling methods for biomolecular simulations. We proposed a new kinetic network model to quantitatively describe replica exchange processes. Using this model, we obtained an analytic efficiency expression for both simulated tempering and replica exchange simulation methods. Our long term aim is to use the developed network models of signaling pathways together with the high resolution QM/MM simulation results to design better initial drug screening strategies for specific inhibitors of kinases with cancer causing mutations.

Dr Rosta is a member of the Thomas Young Centre.

More details can be found here:

Research Projects:

  • QM/MM simulations of metal ion catalysed phosphate hydrolysis and transfer.
  • Kinetic network models of molecular signalling pathways.
  • Ab initio method development exploring substituent effects in Marcus theory of electron transfer

Dr Rosta's Research Portal