Δ2用于反应性质预测的机器学习.
Qiyuan Zhao1, Dylan M Anstine2, Olexandr Isayev2
1Davidson School of Chemical Engineering, Purdue University West Lafayette IN 47906 USA bsavoie@purdue.edu.
一个新的 Δ2 学习模型使用低级几何准确地预测了高级激活能量. 这种机器学习方法可以加速化学反应的表征,准确度高,计算成本低.
科学领域:
- 计算化学计算化学
- 机器学习在化学中的应用
- 量子化学 是一个量子化学.
背景情况:
- Δ-学习模型可以加速高层能量计算,但无法预测反应特性.
- 像激活能量这样的反应性质需要高层次的几何和能量评估.
研究的目的:
- 引入一种新的 Δ2 学习模型,从低级别的临界点几何体中预测高级激活能量.
- 能够准确有效地描述化学反应.
主要方法:
- 开发了一个 Δ2 学习模型,利用原子智能特征化.
- 在使用GFN2-xTB和B3LYP-D3/TZVP能量对~167,000个反应的数据集进行了模型训练.
- 在外部数据集上验证了可转移性,并通过高西安-4计算进行了微调.
主要成果:
- 该 Δ2 模型准确地预测了来自低级别几何学的高水平激活能量.
- 证明了低层和高层结构之间的几何偏差的隐式学习.
- 在看不见的反应和外部测试套件上实现了近乎化学准确性.
结论:
- Δ2学习模型为加速化学反应表征提供了一种有效的策略.
- 将机器学习与半经验量子化学相结合,以获得高精度.
- 微调比DFT预测提高了35%的准确性,而计算成本较低.
更多相关视频
13:54A Workflow for Lipid Nanoparticle LNP Formulation Optimization using Designed Mixture-Process Experiments and Self-Validated Ensemble Models SVEM
Published on: August 18, 2023
10:21Author Spotlight: Streamlining Protein Target Prediction and Validation via Molecular Docking and CETSA
Published on: February 23, 2024
相关概念视频
Predicting Reaction Outcomes
Classification of Titrimetric Analysis Based on Reaction Types
Titrations between an acid and a base lead to neutralization reactions that form...
Predicting Products: SN1 vs. SN2
With increased substitution on the alkyl halide,...
Predicting Products: Substitution vs. Elimination
The following factors can influence the mechanisms competing against each other:
Measuring Reaction Rates
Rate Law and Reaction Order
For example, in a generic reaction aA + bB ⟶ products, where a and b are stoichiometric coefficients, the rate law can be written as:
rate = k[A]m[B]n
[A] and [B] represent the molar concentrations of reactants, and k is the rate...
