The Molecular Simulation ToolBox: The best of two Worlds

One of the main methods in computational biophysics is molecular dynamics
simulations, which is mainly based on propagating Newton's equations, to evolve the
system of interest in a given period of time. Its growth has increased since the arrival
of graphics processors (GPUs), becoming an excellent tool that allows
complementing experiments, and, in several cases, generates results that are not
reachable by experimental techniques. However, one of the main limitations of this
method is the need to build an initial system, with the desired conditions, which
implies technical knowledge and laborious procedures. Thus, our objective was to
develop a set of tools in Python that facilitate the assembly of complex systems.
Although other tools currently exist for this task, many of them focus on a
user-friendly environment, which makes it difficult to prepare multiple systems at the
same time. Our code brings together the best of two worlds, connecting
CHARMM-GUI with PSFGen, allowing us to overcome the limitations of the size of
systems to be created imposed by CHARMM-GUI, and using the versatility of
PSFGen. As a case study, we present different systems, which include protein in
solution, protein in membrane, protein-ligand complexes and a big system that is the
dengue viral cap in solution with around 17 million atoms, which our code
successfully built in just 2 hours, compared to the significantly longer time required
by CHARMM-GUI for the same task. In conclusion, our code is optimized to quickly
prepare different systems, by combining CHARMM-GUI with PSFGen, which
provides a framework to assemble systems at a cellular scale.