Login / Signup

FixBox: A General Algorithm to Fix Molecular Systems in Periodic Boxes.

António M BaptistaLucie da RochaSara R R Campos
Published in: Journal of chemical information and modeling (2022)
Periodic boundary conditions (PBCs) are a standard feature of molecular simulations, and their mathematical and computational aspects are well-understood and relatively straightforward. However, they can in practice be a nuisance when simulating heterogeneous systems, especially when different types of molecules change their relative positions during the simulation. Although the translation required to fix a broken molecular complex of interest can in most cases be easily inferred by visual inspection, it typically depends on the type of system, its configuration, and the box geometry, making automated procedures problematic. We present here a general algorithm, named FixBox, that can fix a molecular complex of interest from a minimal set of definitions of its assembling parts and intended arrangement in the simulation box. It uses a unified triclinic framework for the box geometric periodicity, does not require a full molecular topology, and is applicable to various types of systems and configurations, making it possible to fully and easily automate the fixing of a broken molecular complex. The performance of the algorithm is illustrated with problematic configurations of various types of simulated systems. The presented formal framework can generally be useful for algorithms that need to perform geometrical transformations on systems with PBCs.
Keyphrases
  • machine learning
  • deep learning
  • single molecule
  • transcription factor
  • primary care
  • binding protein
  • quality improvement
  • virtual reality