Combining Simulations, Theory, and Experiments into Multiscale Models of Biological Events

The field of computational biophysics has recently undertaken a shift toward molecular simulations performed in cell-mimicking conditions, coming as far as to simulate a portion of the bacterial cytoplasm. To model even larger and more complex systems, for timescales longer than the millisecond, the atomistic structures may be simplified to coarse-grained representations (e.g. one pseudo-atom per amino acid). In principle, if the system is increasingly simplified (e.g. one pseudo-atom per protein or even using a continuum description), slower and slower molecular motions become accessible. This multi-scale approach, while extremely powerful, poses several challenges in practice. One such challenge is to obtain accurate forces among the pseudo-atoms (force field) considering that these forces must balance several physicochemical effects. It is also challenging to have an accurate conformational ensemble and, simultaneously, the energies of its conformations. This is because the structural simplifications eliminate an entropic contribution. To face the above challenges, and consequently obtain mechanistic insights into biological processes, it is becoming common to build multi-scale models and integrate them with data coming from diverse experimental sources.

In this Research Topic we will discuss recent advances in the field of multiscale modelling and simulation of biological processes, and the insights gained from them. Special attention will be devoted to those approaches that emphasize their integration with experimental data. We will address several challenges in this field. The first one is the choice of a set of reference quantities (experimentally measured or from other simulations) that will be used for the parameterization of a reliable force field or of the equations governing the evolution of the biological process. The second challenge focuses on the reduction of the system’s resolution. Often this reduction is performed using physicochemical intuition, less frequently with automatic methods that try to maximize consistency with experiments. The third challenge is to define the initial “conditions” from which the computation will start. While for a single-domain protein this is relatively straightforward if the protein structure is known, setting up a molecular simulation of a virtual cell would require the modelling of complex supramolecular structures (e.g. the cytoskeleton), as the first step. The fourth challenge is to switch back from a coarse to a finer description, which can be employed to validate the coarse model or to seed higher-resolution simulations.

The multi-scale modeling and simulation of biological processes can largely benefit from the integration with experiments. Several of these techniques, such as cryo-EM, XL-MS, FRET, NMR, SHAPE and super-resolution microscopy, can now access conformational rearrangements and inter-molecular interactions in vivo, on diverse time and length scales. This makes such experimental data ideal ingredients to model a biological process on multiple scales. To underline the potential of this field and strive for communication and collaboration between the experimental and computational communities, we aim at collecting original research papers, reviews and opinions on the following topics:

• Theoretical advances and challenges of integrating experimental data and molecular simulations
• Integrative modeling
• Towards the modeling of eukaryotic cells
• DNA-protein interactions
• Multiscale models of biological membranes
• Recent advances in crosslinking-mass spectroscopy applied to the modeling and simulations of biomolecules
• Multi-scale models and simulations of biofilm growth informed by biological assays
• Cryo-EM and super resolution microscopy: modeling the sub-cellular scale
• DNA nano-structures and DNA origami
• Simulation of the nucleosome and of the genetic material
• Macromolecular crowding and how it affects macromolecule stability and cell function
• Single-molecule FRET experiments integrated into simulations of biomolecules
• Molecular simulations of viruses
• Software for visualizing and building complex biological environments
• Development of coarse-grained force fields: accuracy and transferability
• Bridging molecular model resolutions: strategies, advances and software development
• Mathematical modeling of complex biological events
• Membrane proteins
• Protein aggregation and disease
• Agent-based modeling in biology
• Integration between deep mutational scanning and molecular models
• Integration between NMR and simulations
• Molecular Dynamics with SHAPE
• Multi-scale models of organelles