Tutorial on quantifying and sampling biomolecular ensembles with ShapeGMM.

Area of Science:

• Computational chemistry
• Biophysics
• Statistical mechanics

Background:

• Understanding biomolecular conformations is crucial for drug discovery and protein function.
• Current methods for enhanced sampling and clustering can be computationally intensive.
• Probabilistic modeling offers a promising avenue for analyzing complex molecular dynamics.

Purpose of the Study:

• To introduce a detailed workflow for clustering and enhanced sampling of biomolecular conformations.
• To demonstrate the application of the ShapeGMM methodology for modeling free energy landscapes.
• To refine conformational models using biased sampling techniques.

Main Methods:

• Utilizing the ShapeGMM (Gaussian Mixture Model) methodology for probabilistic modeling of conformations.
• Generating and fitting equilibrium molecular dynamics simulation data.
• Employing Metadynamics with a size-and-shape PLUMED module for enhanced sampling along a reaction coordinate.
• Clustering biased conformations to refine the equilibrium ShapeGMM model.

Main Results:

• Successfully generated and fitted equilibrium molecular dynamics data using ShapeGMM.
• Developed a reaction coordinate between two states using the ShapeGMM model.
• Demonstrated effective enhanced sampling along the reaction coordinate via Metadynamics.
• Achieved refined equilibrium conformational models through clustering of biased samples.

Conclusions:

• The ShapeGMM methodology provides a robust framework for analyzing biomolecular conformations.
• The presented workflow enables efficient enhanced sampling and accurate conformational clustering.
• This approach facilitates a deeper understanding of free energy landscapes and molecular dynamics.