GCIceNet: A Graph Convolutional Network for Accurate Classification of Water Phases

Abstract

Understanding phases of water molecules based on local structure is essential for understanding their anomalous properties. However, due to complicated structural motifs formed via hydrogen bonds, conventional order parameters represent the water molecules incompletely. In this paper, we develop a GCIceNet, which automatically generates machine-based order parameters for classifying the phases of the water molecules via supervised and unsupervised learning. Multiple graph convolutional layers in the GCIceNet can learn topological informations of the complex hydrogen bond networks. It shows a substantial improvement of accuracy for predicting the phase of water molecules in the bulk system and the ice/vapor interface system. A relative importance analysis shows that the GCIceNet can capture the structural features of the given system hidden in the input data. Augmented with the vast amount of data provided by molecular dynamics simulations, the GCIceNet is expected to serve as a powerful tool for the fields of glassy liquids and hydration layers around biomolecules.

References in corpus (13)

• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• Quantum-Chemical Insights from Deep Tensor Neural Networks
• Machine learning phases of matter
• Accurate determination of crystal structures based on averaged local bond order parameters
• Understanding water's anomalies with locally favored structures
• Direct Calculation of Ice Homogeneous Nucleation Rate for a Molecular Model of Water
• Advances in the experimental exploration of water's phase diagram
• Homogeneous ice nucleation evaluated for several water models
• How Do Quasicrystals Grow?
• Mixture-like behavior near a liquid-liquid phase transition in simulations of supercooled water
• Quantification of Order in the Lennard-Jones System
• Unsupervised learning for local structure detection in colloidal systems
• Searching for crystal-ice domains in amorphous ices

Cited by in corpus (3)

• The Seven Deadly Sins: when computing crystal nucleation rates, the devil is in the details
• Recent advances in describing and driving crystal nucleation using machine learning and artificial intelligence
• Controlling crystallization: What liquid structure and dynamics reveal about crystal nucleation mechanisms