The roots of the molecular simulation application can be traced back to physics where it was applied to simplified hard-sphere systems [1]. This field of molecular simulation study has gained a lot of interest since then and applied to perform simulations to fold small protein at multi-microsecond scale [2-4], predict functional properties of receptors and to capture the intermediate transitions of the complex [5], and to study the movement and behavior of ligand in a binding pocket and also to predict interactions between receptors and ligands [6,7].

GROMACS is the most widely used software implemented to study the molecular dynamics (MD) simulations of complex proteins [8]. GROMACS offers a set of commands which can be easily executed for MD simulation of a protein or to a complex protein with a ligand to study protein folding kinetics to computational drug design to the refinement of molecular structures. Recently,

Irrgang et al., [9] have proposed an API for GROMACS called “gmxapi” written in pure Python and implemented as a C++ extension.

This API allows the users to simply construct the computational task graphs permitting the parallel optimizations and mixing of MD simulation and machine-learning operations using other software packages such as TensorFlow [10]. The API provides a native interface to GROMACS MD engine [11]. Users can simply drive MD simulations via high-level procedural commands, an object-oriented interface, or can employ their own extension code.

The restrained-ensemble simulations compute population properties from a set of MD simulation data, then compare these computed simulations to residue-residue distance distributions used as experimental data measured via double electron-electron resonance (DEER) spectroscopy. Then, a distance histogram is calculated by the simulation algorithm from the estimated ensemble and calculates a distance-dependent biasing force for the simulations, which are run for an interval of time (Δt) before repeating the process [9].

gmxapi enables custom plugins for user-defined forces, allows custom potential functions, provides the optimized performance of the software GROMACS, and allows to build and execute computational graphs.

References

  1. Alder, B. and Wainwright, T. (1957) Phase transition for a hard sphere system. J. Chem. Phys., 27, 1208–1209.
  2. van der Spoel, D. and van Maaren, P.J. (2006) The origin of layer structure artifacts in simulations of liquid water. J. Chem. Theor. Comput., 2, 1–11.
  3. Lindorff-Larsen, K. et al. (2011) How fast-folding proteins fold. Science, 334, 517–520.
  4. Bowman, G. et al. (2011) Atomistic folding simulations of the five helix bundle protein 6-85. J. Am. Chem. Soc., 133, 664–667.
  5. Nury, H. et al. (2010) One-microsecond molecular dynamics simulation of channel gating in a nicotinic receptor homologue. Proc. Natl Acad. Sci. USA, 107, 6275–6280.
  6. Chong, L. et al. (1999) Molecular dynamics and free-energy calculations applied to affinity maturation in antibody 48g7. Proc. Natl Acad. Sci. USA, 96, 14330–14335.
  7. Huang, D. and Caflisch, A. (2011) The free energy landscape of small molecule unbinding. PLoS Comput. Biol., 7, e1002002.
  8. Hess, B. et al. (2008) Gromacs 4: Algorithms for highly efficient, load-balanced, and scalable molecular simulation. J. Chem. Theor. Comput., 4, 435–447.
  9. M Eric Irrgang, Jennifer M Hays, Peter M Kasson (2018) gmxapi: a high-level interface for advanced control and extension of molecular dynamics simulations, Bioinformatics, bty484,
  10. https://www.tensorflow.org/
  11. Pronk, S., et al. GROMACS 4.5: a high-throughput and highly parallel open source molecular simulation toolkit. Bioinformatics 2013;29(7):845-854