CoCo-MD: A Simple and Effective Method for the Enhanced Sampling of Conformational Space

Area of Science:

• Computational Chemistry
• Biophysics
• Molecular Dynamics

Background:

• Principal Component Analysis (PCA) is a dimensionality reduction technique.
• Ensemble enrichment methods aim to improve the sampling of molecular configurations.
• CoCo (complementary coordinates) was initially developed for NMR data analysis.

Purpose of the Study:

• To evaluate the potential of the CoCo method combined with molecular dynamics (CoCo-MD) for enhanced conformational sampling.
• To investigate CoCo-MD's general applicability beyond NMR data analysis.
• To develop efficient workflows for free energy calculations and data unbiasing.

Main Methods:

• Iterative workflow: interleaving short multiple-walker MD simulations with CoCo-informed long-range jumps.
• Utilizing reservoir-REMD for free energy calculations.
• Developing an approximate, fast, and practical method for unbiasing CoCo-MD data.

Main Results:

• Achieved up to a 10-fold increase in conformational sampling rate for the alanine penta-peptide model system.
• Demonstrated significantly enhanced conformational sampling for cyclosporine A with reduced computational resources.
• Successfully simulated the transition from the open to the closed state of the maltose binding protein.

Conclusions:

• CoCo-MD offers a powerful and general approach for enhanced sampling of conformational space.
• The PCA-based strategy automatically identifies adaptive collective variables, eliminating the need for user pre-definition.
• The method is compatible with standard MD packages and requires no code modification.