Learning Collective Variables with Synthetic Data Augmentation through Physics-Inspired Geodesic Interpolation

Abstract

In molecular dynamics simulations, rare events, such as protein folding, are typically studied using enhanced sampling techniques, most of which are based on the definition of a collective variable (CV) along which acceleration occurs. Obtaining an expressive CV is crucial, but often hindered by the lack of information about the particular event, e.g., the transition from unfolded to folded conformation. We propose a simulation-free data augmentation strategy using physics-inspired metrics to generate geodesic interpolations resembling protein folding transitions, thereby improving sampling efficiency without true transition state samples. This new data can be used to improve the accuracy of classifier-based methods. Alternatively, a regression-based learning scheme for CV models can be adopted by leveraging the interpolation progress parameter.

References in corpus (15)

• Escaping free-energy minima
• PLUMED 2: New feathers for an old bird
• Well-Tempered Metadynamics: A Smoothly Converging and Tunable Free-Energy Method
• Reversible Work Transiton State Theory: Application to Dissociative Adsorption of Hydrogen
• Statistically optimal analysis of samples from multiple equilibrium states
• Identification of slow molecular order parameters for Markov model construction
• Enhanced sampling methods for molecular dynamics simulations
• Data-Driven Collective Variables for Enhanced Sampling
• A unified framework for machine learning collective variables for enhanced sampling simulations: $\text{mlcolvar}$
• Rare Event Kinetics from Adaptive Bias Enhanced Sampling
• From enhanced sampling to reaction profiles
• Computing the Committor with the Committor: an Anatomy of the Transition State Ensemble
• Deep Learning Collective Variables from Transition Path Ensemble
• Structure and polymerization of liquid sulfur across the $\lambda$-transition
• Is the Local Ion Density Sufficient to Drive NaCl Nucleation from the Melt and Aqueous Solution?

Cited by in corpus (2)

• Enhanced Sampling in the Age of Machine Learning: Algorithms and Applications
• Metatensor and metatomic: foundational libraries for interoperable atomistic machine learning