Wielding the power of interactive molecular simulations

Abstract

Since the dawn of the computer age, scientists have designed devices to represent molecular structures and developed tools to simulate their dynamic behavior in silico. To this day, these tools remain central to our understanding of biomolecular phenomena. In contrast to other fields such as fluid mechanics or meteorology, the observation of molecular motions at the atomic level remains a major experimental challenge. Continuous advances in computer graphics and numerical computation, combined with the emergence of humancomputer interaction approaches, led to the methodology of so-called "interactive molecular simulations", characterized by two main features. First, the possibility to visualize a running simulation in interactive time, that is, compatible with human perception. Second, the possibility to manipulate the simulation interactively by imposing a force, changing a biophysical property, or editing runtime parameters on the fly. Such simulations are still little used in computational biology, where it is more common to run a series of offline simulations and then visualize and analyze the results. However, interactive molecular simulation tools promise to handle time-consuming tasks such as the modeling of particularly complex biomolecular structures more efficiently or to support approaches such as Rational Drug Design with regard to pharmaceutical applications.

Building Biological Relevance Into Integrative Modelling of Macromolecular Assemblies

“…Since integrative modelling often means seeking the best compromise from different and sometimes conflicting sources of information about a given target, the pipeline we have developed explicitly acknowledges that the human operator in the modelling loop may be relied upon to make informed decisions in accordance with the biological data and the physical laws at play in the chosen modelling approach, e.g., in molecular dynamics simulation, constrained minimization, etc. ( Lanrezac et al, 2021 ).…”

All-Atom Biomolecular Simulation in the Exascale Era

“…Another conclusion from the workshop is the pressing need for greater software portability, data uniformity and ready access to the research community, and modern visualization and analysis tools. 22,51 To some extent, the community, like several others, has developed at a rapid pace, with many projects moving in parallel and often independently. This is a natural progression, but with the appearance of several exascale centers around the world, scientific output would be significantly enhanced by more coordinated collective efforts reminiscent of the Linux operating system development.…”

“…12 In addition, several speakers from Julich, Oak Ridge, and elsewhere, rooted in computer science and software engineering, discussed the core computational challenges and opportunities related to exascale computing in biology. 22,23 Other recent reviews and perspectives 24−27 have discussed extensively the current status of biomolecular simulation and modeling, including coarse-grained methods. 27 Here, we focus on atomic-level simulations and both the enormous potential and the resulting challenges created by the emergence of exascale computing.…”

Fast and Interactive Positioning of Proteins within Membranes

“…The relationship between these three software components is shown schematically in Figure 3 and uses standardized protocol and data structures for information exchange. More details on such interactive simulations can be found in [12]. The visualization part uses the Unity game engine for development, whose C# language allows for modularity.…”

“…Computational algorithms range from simple implicit membrane representations [9] to computationally intensive molecular dynamics simulations in fully hydrated lipid bilayers [10,11] and are not interactive. Interactive molecular simulations (IMS), as recently revisited in [12], offer a different approach: the scientist performs the manipulations interactively, providing intuitive insight into the process. Biologically and functionally relevant structural information can be obtained through IMS.…”