Robosample

Robosample applies constrained-dynamics algorithms adapted from robotics to enhance conformational sampling in molecular dynamics (MD) simulations, enabling larger time steps and reduced computational cost.

Key Features:

• Constrained Dynamics Algorithms: Implements Constrained Dynamics Hamiltonian Monte Carlo (CDHMC) as a Gibbs sampling move that permits selective coordinate changes for more efficient exploration of configuration space.
• Diverse Joint Implementations: Supports Cartesian, torsional, spherical, and cylindrical joint dynamics, with spherical and cylindrical joints able to be distributed along molecules.
• Enhanced Sampling Efficiency: Integrates fully flexible moves with torsional, cylindrical, or spherical dynamics to recover free energy surfaces (e.g., alanine dipeptide) and to accelerate Ramachandran transitions by an order of magnitude when only key torsions are mobile.
• Application to Complex Systems: Enables sampling of large biomolecules and complex glycans, covering substantially larger regions of configuration space when combined with constrained dynamics moves.

Scientific Applications:

• Conformational Sampling of Large Biomolecules: Efficiently sample conformational ensembles of large proteins and glycans in MD simulations.
• Free Energy and Transition Studies: Recover free energy surfaces and accelerate conformational transitions, exemplified by alanine dipeptide and Ramachandran dynamics.
• Cost-Effective MD Workflows: Permit larger MD time steps and reduced computational cost through incorporation of constrained-dynamics moves.

Methodology:

Uses CDHMC (Constrained Dynamics Hamiltonian Monte Carlo) implementing Gibbs sampling moves and supports Cartesian, torsional, spherical, and cylindrical joint dynamics, combining fully flexible and constrained moves within MD simulations.