Novel approach for designing order parameters of clathrate hydrate structures by graph neural network.

Clathrate hydrates continue to be the focus of active research efforts due to their use in energy resources, transportation, and storage-related applications. Therefore, it is crucial to define their essential characteristics from a molecular standpoint. Understanding molecular structure in particular is crucial because it aids in understanding the mechanisms that lead to the formation or dissociation of clathrate hydrates. In the past, a wide variety of order parameters have been employed to classify and evaluate hydrate structures. An alternative approach to inventing bespoke order parameters is to apply machine learning techniques to automatically generate effective order parameters. In earlier work, we suggested a method for automatically designing novel parameters for ice and liquid water structures with Graph Neural Networks (GNNs). In this work, we use a GNN to implement our method, which can independently produce feature representations of the molecular structures. By using the TeaNet-type model in our method, it is possible to directly learn the molecular geometry and topology. This enables us to build novel parameters without prior knowledge of suitable order parameters for the structure type, discover structural differences, and classify molecular structures with high accuracy. We use this approach to classify the structures of clathrate hydrate structures: sI, sII, and sH. This innovative approach provides an appealing and highly accurate replacement for the traditional order parameters. Furthermore, our method makes clear the process of automatically designing a universal parameter for liquid water, ice, and clathrate hydrate to analyze their structures and phases.

Graph neural networks classify molecular geometry and design novel order parameters of crystal and liquid.

Molecular dynamics simulation produces three-dimensional data on molecular structures. The classification of molecular structure is an important task. Conventionally, various order parameters are used to classify different structures of liquid and crystal. Recently, machine learning (ML) methods have been proposed based on order parameters to find optimal choices or use them as input features of neural networks. Conventional ML methods still require manual operation, such as calculating the conventional order parameters and manipulating data to impose rotational/translational invariance. Conversely, deep learning models that satisfy invariance are useful because they can automatically learn and classify three-dimensional structural features. However, in addition to the difficulty of making the learned features explainable, deep learning models require information on large structures for highly accurate classification, making it difficult to use the obtained parameters for structural analysis. In this work, we apply two types of graph neural network models, the graph convolutional network (GCN) and the tensor embedded atom network (TeaNet), to classify the structures of Lennard-Jones (LJ) systems and water systems. Both models satisfy invariance, while GCN uses only length information between nodes. TeaNet uses length and orientation information between nodes and edges, allowing it to recognize molecular geometry efficiently. TeaNet achieved a highly accurate classification with an extremely small molecular structure, i.e., when the number of input molecules is 17 for the LJ system and 9 for the water system, the accuracy is 98.9% and 99.8%, respectively. This is an advantage of our method over conventional order parameters and ML methods such as GCN, which require a large molecular structure or the information of wider area neighbors. Furthermore, we verified that TeaNet could build novel order parameters without manual operation. Because TeaNet can recognize extremely small local structures with high accuracy, all structures can be mapped to a low-dimensional parameter space that can explain structural features. TeaNet offers an alternative to conventional order parameters because of its novelty.

Graph-Neural-Network-Based Unsupervised Learning of the Temporal Similarity of Structural Features Observed in Molecular Dynamics Simulations.

Classification of molecular structures is a crucial step in molecular dynamics (MD) simulations to detect various structures and phases within systems. Molecular structures, which are commonly identified using order parameters, were recently identified using machine learning (ML), that is, the ML models acquire structural features using labeled crystals or phases via supervised learning. However, these approaches may not identify unlabeled or unknown structures, such as the imperfect crystal structures observed in nonequilibrium systems and interfaces. In this study, we proposed the use of a novel unsupervised learning framework, denoted temporal self-supervised learning (TSSL), to learn structural features and design their parameters. In TSSL, the ML models learn that the structural similarity is learned via contrastive learning based on minor short-term variations caused by perturbations in MD simulations. This learning framework is applied to a sophisticated architecture of graph neural network models that use bond angle and length data of the neighboring atoms. TSSL successfully classifies water and ice crystals based on high local ordering, and furthermore, it detects imperfect structures typical of interfaces such as the water-ice and ice-vapor interfaces.