ABSTRACT
Classical molecular dynamics (MD) is one of the most widely used biomolecular simulation technique. In the recent years, the increase in the availability of
experimental structural data, the massive improvement in the MD codes and the irruption of new computer platforms (massively parallel supercomputers, GPU-based systems and specific purpose computers) produced a revolution in the field, approaching the possibility to simulate with high quality, biologically relevant systems in relevant time scales. Thus, we have recently seen atomistic simulations of molecular systems containing millions of atoms, others extended to the millisecond scale, or dealing with model systems aiming to reproduce the crowded cellular environments. We have seen major rewriting of computer codes originally developed in the seventies, computers created on purpose to perform MD simulations, massive initiatives of distributed computing and the first projects to manage and store the deluge of data provided by MD simulations. Where are all these new innovations going to drive us? Are we going to see in a close future entire-cell simulations? Are we going to analyze second-long trajectories, or perform real-time MD simulations? In this contribution, we will review the technical limits of current state-of-the-art of MD biosimulations and will try to decipher the future in front of us.
