Statistical Mechanics

One of the main research activities in my group concerns the development and application of advanced theoretical approaches to investigate the structural dynamics of complex macromolecular systems with an emphasis on rare thermally activated conformational reactions of biomolecules. In particular, we have developed an efficient computational scheme denominated Dominant Reaction Pathways, DRP (see papers below). The DRP method is based on combining various kinds of biased molecular dynamics with the path integral formulation of Langevin dynamics. The result is a variational approximation which enables the efficient computation of reaction pathways for proteins consisting of hundreds of amino acids, using atomistic force fields in implicit and explicit solvent. 

 

Atomistic trajectories obtained by means of plain molecular dynamics or through enhanced path sampling techniques contain in principle all the information needed to characterise the reaction mechanism, identify the reaction coordinates, and predict thermodynamical and kinetic observables. However, in practice, extracting this information from the simulation data is usually a formidable effort, due to the high-dimensional space of macromolecular systems. In our group we develop and apply different types schemes aiming to reformulating the dynamics in terms of a relatively small number of descriptors and to achieve low-resolution representations. In this context, Renormalisation Group theory has been employed by our group in order to derive a Langevin dynamics with lower time resolution which can be solved using larger time steps. 

A few representative papers: