Manifold Learning in Atomistic Simulations: A Conceptual Review

Abstract

Analyzing large volumes of high-dimensional data requires dimensionality reduction: finding meaningful low-dimensional structures hidden in their high-dimensional observations. Such practice is needed in atomistic simulations of complex systems where even thousands of degrees of freedom are sampled. An abundance of such data makes gaining insight into a specific physical problem strenuous. Our primary aim in this review is to focus on unsupervised machine learning methods that can be used on simulation data to find a low-dimensional manifold providing a collective and informative characterization of the studied process. Such manifolds can be used for sampling long-timescale processes and free-energy estimation. We describe methods that can work on datasets from standard and enhanced sampling atomistic simulations. Unlike recent reviews on manifold learning for atomistic simulations, we consider only methods that construct low-dimensional manifolds based on Markov transition probabilities between high-dimensional samples. We discuss these techniques from a conceptual point of view, including their underlying theoretical frameworks and possible limitations.

References in corpus (24)

• Well-Tempered Metadynamics: A Smoothly Converging and Tunable Free-Energy Method
• Machine learning for molecular simulation
• Accurate sampling using Langevin dynamics
• Enhanced sampling methods for molecular dynamics simulations
• A Variational Approach to Enhanced Sampling and Free Energy Calculations
• An introduction to the Ginzburg-Landau theory of phase transitions and nonequilibrium patterns
• Variational cross-validation of slow dynamical modes in molecular kinetics
• Machine learning force fields and coarse-grained variables in molecular dynamics: application to materials and biological systems
• Unsupervised machine learning in atomistic simulations, between predictions and understanding
• Optimizing the ensemble for equilibration in broad-histogram Monte Carlo simulations
• Autonomous artificial intelligence discovers mechanisms of molecular self-organization in virtual experiments
• Performance Limitations of Flat Histogram Methods and Optimality of Wang-Landau Sampling
• Unsupervised landmark analysis for jump detection in molecular dynamics simulations
• Computing committors via Mahalanobis diffusion maps with enhanced sampling data
• Visualising energy landscapes through manifold learning
• Spectral Map: Embedding Slow Kinetics in Collective Variables
• Reweighted Manifold Learning of Collective Variables from Enhanced Sampling Simulations
• On UMAP's true loss function
• Stochastic Neighbor Embedding separates well-separated clusters
• Unsupervised learning of representative local atomic arrangements in molecular dynamics data
• Selecting High-Dimensional Representations of Physical Systems by Reweighted Diffusion Maps
• Enhanced Sampling with Machine Learning: A Review
• From $t$-SNE to UMAP with contrastive learning
• T-SNE Is Not Optimized to Reveal Clusters in Data

Cited by in corpus (2)

• Spectral Map for Slow Collective Variables, Markovian Dynamics, and Transition State Ensembles
• Learning Markovian Dynamics with Spectral Maps