DeepModeling

Recently, the research group led by Researcher Xu Shenzhen from the School of Materials Science and Engineering at Peking University collaborated with the Beijing Academy of Intelligent Sciences (AISI) and Deep Potential Technology (DP Technology). They employed the deep potential method[1] to study the nuclear quantum effects in proton-coupled electron transfer during electrochemical reactions and develop computational methods. Notably, all first-principles calculations in this study were carried out using the domestic first-principles software ABACUS [2]. The relevant research findings were published in Nature Communications under the title "Probing Nuclear Quantum Effects in Electrocatalysis via a Machine-Learning Enhanced Grand Canonical Constant Potential Approach" [3]. Doctoral students Sun Menglin, Jin Bin, and Yang Xiaolong are the co-first authors, and Researcher Xu Shenzhen is the corresponding author.

Recently, Associate Professor Hu Zhixin from Tianjin University collaborated with the research groups of Professor Ji Wei and Associate Researcher Wang Cong from Renmin University of China. Based on first-principles calculations and using both VASP and the domestic first-principles software ABACUS, they revealed the microscopic mechanisms of the changes in the easy magnetization axis and topological properties with the number of layers in MnSe₂. The relevant research results were published in the journal Physical Review B under the title "Interlayer coupling driven rotation of the magnetic easy axis in MnSe₂ monolayers and bilayers" (DOI: 10.1103/PhysRevB.111.054422). The first authors of the paper are Zhang Zhongqin and Wang Cong.

Abstract

Recently, Zhao Haotian, a PhD candidate at the University of Science and Technology of China, and Professor He Lixin proposed and implemented a Hybrid Gauge Real-Time Time-Dependent Density Functional Theory (Hybrid gauge rt-TDDFT) applicable to atomic orbital basis sets in the domestic open-source density functional theory software ABACUS. Based on the traditional velocity gauge, this method introduces a time-varying phase dependent on the vector potential, effectively overcoming the systematic errors caused by the incompleteness of the local basis set and providing consistent and reliable simulation results in both periodic and non-periodic systems. At the same time, this method significantly improves the calculation efficiency in periodic systems, offering a new solution that combines accuracy and efficiency for first-principles real-time dynamics simulations under the action of an electric field.

Enzymes are the core catalysts of life activities, and the prediction of their functions has important applications in various fields. However, traditional methods of enzyme classification, such as EC numbers or protein family classification, rely on manual classification, which can lead to inaccurate classification granularity. Moreover, these methods lack the ability to characterize the dynamic structural transformation of substrates and products during the reaction process, making it difficult to accurately understand the actual functions and catalytic mechanisms of enzymes. Against this backdrop, ReactZyme, developed by Hua et al., brings new hope to the study of enzyme functions. Uni - Mol, as the core engine for molecular modeling, provides powerful support for capturing the complex three - dimensional interactions between enzymes and substrates. This achievement was presented at the Datasets and Benchmarks track of the 38th Conference on Neural Information Processing Systems (NeurIPS 2024), and its code and data have been open - sourced(https://github.com/WillHua127/ReactZyme).

The machine learning potential method provides a powerful means for high-precision atomic-scale simulations and has profoundly influenced the research paradigm in the materials field. Recently, some magnetic model methods have incorporated magnetic physical quantities such as magnetic moments into machine learning potentials, offering new tools for the study of magnetic materials. However, efficient models rely on accurate data, and currently, the development of magnetic models lacks efficient and accurate data generation tools.

Recently, researchers from the AI for Science Institute (AISI), the Graduate School of the China Academy of Engineering Physics, and their collaborators pre-published an article titled "Integrating Deep-Learning-Based Magnetic Model and Non-Collinear Spin-Constrained Method: Methodology, Implementation and Application" on arXiv [1]. Based on the domestic open-source density functional software ABACUS, the article implemented a constrained density functional method, DeltaSPIN, for arbitrary non-collinear magnetic moments. They trained a magnetic model of the pure iron system using ABACUS + DeePSPIN and accurately simulated the ferromagnetic-paramagnetic transition of the BCC phase.

The research team led by Associate Professor Kun Cao from the School of Physics, Sun Yat-sen University, has made significant progress in the study of the magnetism of bilayer nickel-based superconductor La₃Ni₂O₇₋_δ under normal temperature and pressure. Through density functional theory (DFT) calculations and Monte Carlo simulations, this research has for the first time revealed the regulatory mechanism of oxygen vacancies on the magnetic ground state and phase transition temperature (T_SDW) of the material, providing a new perspective for understanding the relationship between high-temperature superconductivity and magnetism in nickel oxides.

This achievement was published in the journal npj Quantum Materials (Paper link: https://www.nature.com/articles/s41535-025-00740-z).

In current biological and chemical research, multimodal learning has become a crucial tool for advancing drug discovery and molecular understanding. However, most existing pre-training frameworks are limited to two modalities, and designing a unified network capable of handling multiple modalities such as natural language, 2D molecular graphs, 3D molecular conformations, and 3D proteins remains challenging. To address this issue, researchers from Pennsylvania State University and Tsinghua University proposed MolBind, a framework that trains multimodal encoders through contrastive learning and maps all modalities to a shared feature space. Uni-Mol plays a key role in this paper. As the core model for 3D conformation encoders and 3D protein encoders, it can effectively capture the 3D structural information of molecules and proteins, providing important support for MolBind to achieve accurate multimodal alignment. The preprint of the research "MOLBIND: Multimodal Alignment of Language, Molecules, and Proteins" was published on arXiv (https://arxiv.org/pdf/2403.08167).

Since the release of the DeePTB v1 version[1], which supports the prediction of Slater-Koster tight-binding (TB) parameters, in Nature Communications in 2023, the DeePTB project has achieved a series of breakthroughs. Subsequently, the DeePTB team introduced the E3 equivariant version[2]. Its core method, the E3 equivariant representation based on Strictly Localized Equivariant Message-passing (SLEM), has made a breakthrough in the processing of quantum operator matrices, enabling more accurate and efficient representation of key quantum operators such as the Kohn-Sham (KS) Hamiltonian, density matrix, and overlap integral. The related research has been accepted as a Spotlight article at the ICLR 2025 conference. On this basis, the team further combined the DeePTB method with the Non-equilibrium Green's Function (NEGF) method to introduce the AI-accelerated quantum transport simulation framework DeePTB-NEGF[3]. This framework significantly improves the computational efficiency of quantum transport simulations. It has been successfully applied to transport calculations in high-throughput and large-scale scenarios, providing a new and efficient tool for quantum transport theoretical research and semiconductor device simulations.

The research group led by Shi Liu from the Department of Physics, School of Science, Westlake University has made the latest progress in revealing the mechanism by which interstitial doping reduces the coercive field of ferroelectric hafnium oxide. On February 4, the research results were published in Physical Review Letters under the title "Origin of Interstitial Doping Induced Coercive Field Reduction in Ferroelectric Hafnia".

In this study, the research group led by Shi Liu comprehensively used first-principles calculations and deep potential-based molecular dynamics to reveal the relationship between interstitial doping and the coercive field of ferroelectric hafnium oxide (HfO₂). It was found that interstitial hafnium doping significantly reduces the energy difference between the polar orthorhombic O phase (space group: Pca2₁) and the tetragonal T phase, thereby reducing the polarization reversal energy barrier and the coercive field. Compared with the theoretical model proposed in previous studies, which suggested that doping-induced rhombohedral R phase reduces the coercive field, this study proposed that the low-doped O phase can better explain the experimental observations of interstitial-doped hafnium-based thin films. Large-scale molecular dynamics simulations based on deep potential further indicated that interstitial hafnium doping can induce the formation of mobile Pbcn-type domain walls, thus reducing the polarization reversal electric field to less than 1 MV/cm. In addition, first-principles calculations revealed a negative correlation between the polarization reversal energy barrier and the radius of the doped atom, and several potential interstitial doping atoms that can effectively reduce the coercive field were screened out.

In machine learning potential development, it is challenging to sample complex chemical spaces with a small amount of data. The research group led by Dongping Chen from Beijing Institute of Technology constructed a feature dataset with low redundancy and low data requirements based on a data dimensionality reduction method. This dataset was combined with the DeePMD method to develop a machine learning potential with high accuracy, high efficiency, and wide application conditions, aiming to describe the interactions in aluminum-lithium (Al-Li) alloys and ammonium perchlorate (AP) interfaces. The relevant research results were published in the journal Journal of Chemical Theory and Computation under the title "Minimizing Redundancy and Data Requirements of Machine Learning Potential: A Case Study in Interface Combustion" (DOI: 10.1021/acs.jctc.4c00587)[1]. Xiaoya Chang, a doctoral student, is the first author.