Séminaire par Mirco Zerbetto, Groupe de Chimie Théorique, Université de Padova, Italie
Dans le cadre de l’animation scientifique, Mirco Zerbetto, chercheur dans le groupe de Chimie Théorique à l’Université de Padova, présentera une conférence intitulée “Stochastic modeling of flexible macromolecules for the interpretation of magnetic spectroscopies” le mardi 10 septembre à 11h en salle de séminaire de l’ISA.
Résumé de la présentation:
Dynamics of complex systems, such as macromolecules or supramolecular assemblies, can be of primary importance in determining their physico-chemical properties. Substrate recognition and allosteric effects in enzymes, channels opening in membranes, dissipative self-assembling systems are just few examples of molecular processes that are activated and regulated by the system dynamics.
Molecular dynamics (MD) simulations are computer experiments that allow to access information on the time evolution of a molecular system, from which dynamic observables can be estimated and employed to interpret experimental observations. For example, orientational spectral densities of a nuclear magnetic probe obtained via MD simulations shall be employed to compute NMR relaxation data. However, in complex systems the molecular motions range from sub-picosecond local dynamics of atoms or groups of atoms to the nanosecond (and up) timescales characteristic of large amplitude motions of entire molecular domains. The latter, slow motions are those that are usually of interest. From the point of view of an MD simulation, slow, collective motions enter in the definition of rare events, i.e. events that occur rarely because large Gibbs free energy barriers need to be crossed. At present, achieving MD trajectories sufficiently long to describe correctly the statistics of such rare events for complex systems is computationally unaffordable.
Multiscale approaches are a useful modeling strategy to deal with the slow dynamics of complex systems. The basic philosophy is to divide the problem into different parts treated at different levels of accuracy. The challenge is to make the computational protocol self-consistent and capable to estimate the wanted observable without resorting to any fitting procedure to free parameters. A multiscale approach will be discussed, in which the molecular degrees of freedom (d.o.f.) are divided into two sets using time-scale separation arguments: the relevant set of d.o.f., directly affecting the observable of interest (e.g., a spectral density), and the irrelevant set of d.o.f. [1,2]. The latter shall influence the physical observable due to the coupling to the relevant degrees of freedom. Short (a few ns) standard or biased MD simulations are employed to evaluate the dependence of the free energy surfaces from relevant d.o.f. Hydrodynamic methods are then employed to evaluate dissipative properties of the system. Finally, a Fokker-Planck equation, from which the irrelevant d.o.f. have been projected out, is employed to describe a stochastic, long-time dynamics in the relevant d.o.f. only. New perspectives will be introduced about a generalized protocol for macromolecules with many degrees of freedom obtained from rigorous, first-principles treatment.
Examples will be shown on the application of the approach to interpret NMR relaxation data of systems with many degrees of freedom [3,4], stressing how the thorough setup of a multiscale computational protocol can lead to the interpretation of observables of complex systems in an ab initio fashion.
- Polimeno, A.; Zerbetto, M.; Abergel, D. J. Chem. Phys. 2019, 150, 184107.
- Polimeno, A.; Zerbetto, M.; Abergel, D. J. Chem. Phys. 2019, 150, 184108.
- Piserchia, A.; Zerbetto, M.; Salvia, M.-V.; Salassa, G.; Gabrielli, L.; Mancin, F.; Rastrelli, F.; Frezzato, D. J. Phys. Chem. C 2015, 119, 20100-21010.
- Rampino, S.; Zerbetto, M.; Polimeno, A. Molecules 2021, 26, 2418.
