Machine learning evaluation of structural descriptors for supercooled water
arXiv:2605.00415 · doi:10.1038/s42004-026-02097-1
Abstract
The anomalous behavior of liquid water is widely associated with a liquid-liquid phase transition between high- and low-density states in the supercooled regime. At the microscopic level, tetrahedral hydrogen-bond networks govern these properties, motivating structural descriptors that characterize local molecular environments. These structural descriptors quantify features such as tetrahedral order, local density, and the separation between the first and second coordination shells; however, they have largely been proposed independently, with limited systematic comparison. Here we evaluate 16 previously proposed descriptors using a neural-network-based temperature classification framework, enabling an objective assessment of their ability to distinguish temperature-dependent structural changes in supercooled water. We further apply an explainable artificial intelligence method that identifies the structural features responsible for the model predictions. This approach reveals how different descriptors encode local structural information and establishes a data-driven framework for benchmarking structural descriptors in liquid water.
15 pages, 5 figures, 1table for main text, 10 figures for supplementary information
References in corpus (22)
- Relation Between the Widom line and the Strong-Fragile Dynamic Crossover in Systems with a Liquid-Liquid Phase Transition
- Understanding water's anomalies with locally favored structures
- Entropy-driven liquid-liquid separation in supercooled water
- Dynamics of simulated water under pressure
- Thermodynamics of supercooled water
- Two-State Thermodynamics and the Possibility of a Liquid-Liquid Phase Transition in Supercooled TIP4P/2005 Water
- Mixture-like behavior near a liquid-liquid phase transition in simulations of supercooled water
- Two-structure thermodynamics for the TIP4P/2005 model of water covering supercooled and deeply stretched regions
- Autonomously revealing hidden local structures in supercooled liquids
- Structure of the First and Second Neighbor Shells of Water: Quantitative Relation with Translational and Orientational Order
- Identifying time scales for violation/preservation of Stokes-Einstein relation in supercooled water
- Unsupervised learning for local structure detection in colloidal systems
- Unravelling the contribution of local structures to the anomalies of water: the synergistic action of several factors
- Crucial role of fragmented and isolated defects in persistent relaxation of deeply supercooled water
- A local order metric for condensed phase environments
- Relation between the Liquid-Liquid Phase Transition and Dynamic Behavior in the Jagla Model
- Structural and topological changes across the liquid-liquid transition in water
- Do Machine-Learning Atomic Descriptors and Order Parameters Tell the Same Story? The Case of Liquid Water
- Beyond Local Structures In Critical Supercooled Water Through Unsupervised Learning
- Identification of local structures in water from supercooled to ambient conditions
- Graph neural network-based structural classification of glass-forming liquids and its interpretation via self-attention mechanism
- Dynamical Heterogeneity in Supercooled Water and its Spectroscopic Fingerprints