Drug design: Insights from atomistic simulations

Computer simulations have become a widely used and powerful tool to study the behaviour of many-particle and many-interaction systems and processes such as nucleic acid dynamics, drug-DNA interactions, enzymatic processes, membrane, antibiotics. The increased reliability of computational techniques has made possible to plane a bottom-up approach in drug design, i.e. designing molecules with improved properties starting from the knowledge of the molecular mechanisms. However, the in silico techniques have to face the fact that the number of degrees of freedom involved in biological systems is very large while the time scale of several biological processes is not accessible to standard simulations. Algorithms and methods have been developed and are still under construction to bridge these gaps. Here we review the activities of our group focussed on the time-scale bottleneck and, in particular, on the use of the metadynamics scheme that allows the investigation of rare events in reasonable computer time without reducing the accuracy of the calculation. In particular, we have devoted particular attention to the characterization at microscopic level of translocation of antibiotics through membrane pores and ligand-receptor interactions, aiming at the identification of structural and dynamical features helpful for a rational drug design.