多体功能纠正的神经网络与原子注意力 (MBNN-att) 用于分子性质预测
Zheng-Xin Yang1, Xin-Tian Xie1, Pei-Lin Kang1
1Collaborative Innovation Center of Chemistry for Energy Material, Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials, Key Laboratory of Computational Physical Science, Department of Chemistry, Fudan University, Shanghai 200433, China.
Journal of chemical theory and computation
|July 22, 2024
概括
一个新的机器学习模型,MBNN-att,可以在CPU上高精度地预测分子性质. 这种模型将原子注意力机制集成到多体神经网络中,以进行高效的化学预测.
科学领域:
- 计算化学的计算化学
- 机器学习在化学中的应用
背景情况:
- 机器学习 (ML) 模型越来越多地用于化学性质预测.
- 现有的模型往往缺乏在中央处理器 (CPU) 设备上可访问的通用,高性能功能.
- 对于广泛的化学应用,存在需要低成本,高效的ML模型.
研究的目的:
- 引入一种新的ML模型,MBNN-att,用于预测分子和材料特性.
- 在CPU设备上实现高性能和通用性.
- 提高化学ML模型的准确性和可转移性.
主要方法:
- 通过将原子注意力机制纳入多体神经网络 (MBNN) 开发了MBNN-att模型.
- 使用显式函数描述符作为基于原子的前神经网络 (NN) 的输入.
- 实现了一个多头自我注意力机制,使用NN的矢量输出,将其分为原子注意力重量和多体函数.
主要成果:
- 在所有QM9属性上,MBNN-att表现出色,其误差低于化学准确度.
- 该模型特别擅长预测与能源相关的广泛性质.
- 系统的比较证实了MBNN-att在其他基于描述器和图形表示的ML模型上的优势.
结论:
- 多体功能框架和原子注意力机制对于MBNN-att的高性能至关重要.
- 在分子性质预测中,MBNN-att表现出强大的可转移性.
- 开发的模型为计算化学中的可访问,高性能ML提供了一个有希望的解决方案.
相关概念视频
02:27Predicting Molecular Geometry
34.2K
VSEPR Theory for Determination of Electron Pair Geometries
34.2K
02:35Noncovalent Attractions in Biomolecules
17.4K
17.4K
02:00Molecular Models
38.2K
Physical models representing molecular architectures of chemical compounds play essential roles in understanding chemistry. The use of molecular models makes it easier to visualize the structures and shapes of atoms and molecules.
38.2K
01:07Nuclear Overhauser Enhancement (NOE)
655
Irradiation of a spin-active nucleus causes an increase or decrease in the signal intensity of neighboring nuclei that are not necessarily chemically bonded or involved in J-coupling. This phenomenon, called the Nuclear Overhauser Enhancement (NOE), results from through-space interactions between the nuclear spins. The NOE effect decreases with increasing internuclear distance and is generally not observed beyond 4 angstroms. In NOE, dipole-dipole interactions between neighboring...
655
02:25Molecules with Multiple Chiral Centers
11.5K
Molecules that possess multiple chiral centers can afford a large number of stereoisomers. For instance, while some molecules like 2-butanol have one chiral center, defined as a tetrahedral carbon atom with four different substituents attached, several molecules like butane-2,3-diol have multiple chiral centers. A simple formula to predict the number of stereoisomers possible for a molecule with n chiral centers is 2n. However, there can be a lower number where some of the stereoisomers are...
11.5K
01:27Atomic Absorption Spectroscopy: Overview
1.8K
Atomic absorption spectroscopy (AAS) is a technique used to analyze elements by measuring electromagnetic radiation (EMR) absorbed by atoms, which causes them to transition to a higher-energy orbit. The most crucial step in AAS is atomization, where the analyte is converted into gas-phase atoms, typically through a flame or furnace. Some of these atoms become thermally excited in the flame, while most remain in the ground state.
When irradiated by EMR of a particular wavelength, these...
When irradiated by EMR of a particular wavelength, these...
1.8K