This technical note describes the recent updates of Graphormer, including architecture design modifications, and the adaption to 3D molecular dynamics simulation. The "Graphormer-V2" could attain better results on large-scale molecular modeling datasets than the vanilla one, and the performance gain could be consistently obtained on downstream tasks. In addition, we show that with a global receptive field and an adaptive aggregation strategy, Graphormer is more powerful than classic message-passing-based GNNs. Graphormer-V2 achieves much less MAE than the vanilla Graphormer on the PCQM4M quantum chemistry dataset used in KDD Cup 2021, where the latter one won the first place in this competition. In the meanwhile, Graphormer-V2 greatly outperforms the competitors in the recent Open Catalyst Challenge, which is a competition track on NeurIPS 2021 workshop, and aims to model the catalyst-adsorbate reaction system with advanced AI models. All models could be found at \url{https://github.com/Microsoft/Graphormer}.
Differentiable Scaffolding Tree for Molecular Optimization
Amplitude(Amplitude公司);Massachusetts Institute of Technology(麻省理工学院);University of Illinois at Urbana-Champaign(伊利诺伊大学厄巴纳-香槟分校);Georgia Institute of Technology(佐治亚理工学院)
Reactive molecular dynamics at constant pressure via non-reactive force fields: extending the Empirical Valence Bond method to the isothermal-isobaric ensemble
State of the art for ab initio vs empirical potentials for predicting 6e$^{-}$ excited state molecular energies: Application to Li$_{2}\left(b,1^{3}Π_{u}\right)$
Accurate non-empirical range-separated hybrid van der Waals density functional for complex molecular problems, solids, and surfaces
我们在非经验vdW-DF方法中引入了一种新的通用范围分离混合范德华密度\ph{functional, termed vdW-DF-ahbr,} [JPCM 32,393001(2020)]。它将vdW-DF 2的相关性与由\ph{a new model of exchange effects}在密度显式vdW-DF 2-b86 r函数中固定的筛选Fock交换相结合[PRB 89,121103(R)(2014)]。新的vdW-DF 2-ahbr防止虚假的交换绑定,并有一个小的密度梯度形式集从多体微扰分析。它是准确的\ph{bulk as well as layered materials},它系统地和显着地提高了目前的vdw-DF的性能的分子问题。重要的是,vdW-DF 2-ahbr也优于目前的标准(色散校正)范围分离的杂交广泛收集的非共价相互作用的基准集,而在同一时间成功地减轻密度驱动的错误,往往影响分子过渡态和异构化计算的描述。vdW-DF 2-ahbr还通过1)正确预测Pt(111)上CO吸附的基底结构和位点偏好,2)在描述CO $_2$在官能化和简单金属-有机骨架中的吸附方面优于现有的非经验vdW-DF,和3)在DNA装配模型中是高度精确的碱基对相互作用的\ph{for the}集合。
Universal Machine-learning Molecular Dynamics at the Speed of Empirical Potentials
Peking University(北京大学);AI for Science Institute(人工智能科学研究所);DP Technology(深势科技);Academy for Advanced Interdisciplinary Studies, Peking University(北京大学交叉信息研究院);Institute of Applied Physics and Computational Mathematics(应用物理与计算数学研究所)
Accelerated Computation of Free Energy Profile at ab Initio Quantum Mechanical/Molecular Mechanics Accuracy via a Semi-Empirical Reference Potential. I. Weighted Thermodynamics Perturbation
Socrates-Mol: Self-Oriented Cognitive Reasoning through Autonomous Trial-and-Error with Empirical-Bayesian Screening for Molecules
State Key Laboratory of Heavy Oil Processing, China University of Petroleum (Beijing), Beijing 102249, China(石油加工国家重点实验室,中国石油大学(北京));Shanghai AI Laboratory,L1 Building, International Media Port, No. 129 Longwen Road, Xuhui District, Shanghai(上海人工智能实验室);College of Energy Innovation, China University of Petroleum (Beijing), Beijing 102249, China(能源创新学院,中国石油大学(北京))
University of Luxembourg(卢森堡大学);Technische Universität Berlin(柏林工业大学);Berlin Institute for the Foundations of Learning and Data – BIFOLD(柏林学习与数据基础研究所(BIFOLD));Max Planck Institute for Informatics(马克斯·普朗克 informatics 研究所);Korea University(高丽大学)
ExoMol line lists -- XL. Ro-vibrational molecular line list for the hydronium ion (H$_3$O$^+$)
计算水合氢(H $_3$ $^{16}$ O $^+$)的新的线列表。该谱线是基于一个新的从头计算偶极矩曲面(CCSD(T)/aug-cc-pVQZ)和一个新的经验势能面(PES)。H $_3$ O $^+$的经验PES是通过对从文献中收集的实验确定的振转能量进行全局拟合来精炼从头算表面而获得的,所述文献包括地面、$ν_1^{\pm}$、$ν_2^{\pm}$、$2ν_2^{\pm}$、$ν_3^{\pm}$和$ν_4^{\pm}$振动态。该线路表涵盖了高达10 000 cm $^{-1}$(波长$>1 μ$ m)的波数范围,并应在高达$T=1500$ K的温度下完成。这是H $_3$ O $^+$的第一个全面的线列表,具有广泛的波数覆盖范围和准确的过渡概率。讨论了太阳系巨行星和系外行星光谱中水合氢离子探测的前景。eXeL细胞系列表可从ExoMol和CDS数据库公开获得。
Development of a Non-Empirical Exchange-Hole Dipole Moment Dispersion Model
University of Toronto(多伦多大学);Acceleration Consortium, University of Toronto(多伦多大学加速联盟);Vector Institute for Artificial Intelligence(向量人工智能研究所);Universidad de Oviedo(奥维耶多大学)