相关实验视频

Updated: Jun 9, 2025

Atomic Scale Structural Studies of Macromolecular Assemblies by Solid-state Nuclear Magnetic Resonance Spectroscopy

Atomic Scale Structural Studies of Macromolecular Assemblies by Solid-state Nuclear Magnetic Resonance Spectroscopy

Published on: September 17, 2017

在¹³C NMR光谱学中使用信息传递神经网络 (MPNNs) 集合进行化学转移预测

D Williamson¹, S Ponte¹, I Iglesias¹

¹Mestrelab Research SL, C/Feliciano Barrera, 9B-Bajo, 15706 Santiago de Compostela, Spain.

Journal of magnetic resonance (San Diego, Calif. : 1997)

|October 31, 2024

概括

使用传递信息的神经网络 (MPNNs) 的深度学习准确地预测了13C NMR的化学变化. 较大,多样化的训练数据集显著提高了对光谱分配的模型概括性和错误估计.

科学领域:

计算化学的计算化学
机器学习机器学习
频谱学是一种光谱学.

背景情况:

核磁共振 (NMR) 光谱对于阐明分子结构至关重要.
预测13C NMR化学转移有助于光谱分配和结构验证.
需要准确的计算方法来补充实验性NMR数据.

研究的目的:

开发和评估一种使用消息传递神经网络 (MPNNs) 的深度学习方法,用于预测13C NMR化学转移.
调查培训数据集大小和多样性对预测准确性和概括性的影响.
建立使用预测的化学转移和相关误差估计来进行光谱分配和验证的定量框架.

主要方法:

利用传递信息的神经网络 (MPNNs) 进行化学转移预测.
在两个数据集上训练模型: ~4,000 和 >40,000 标记小分子结构.
实施了一个整体框架,以减少随机变化和增强强性.
开发了基于训练数据属性和概率评估的错误估计方法.

主要成果:

在更大,更多样化的数据集上训练的MPNN合集显示出对未见数据的优越通用性.
预测性能在数据集内测试集上是可比的,但在交叉数据集和外部测试集上有显著差异.
较大的数据集可以实现更强大的错误建模,将预测错误与特征向量频率相关联.
一个高斯核的累积分布准确地建模了结构级预测错误,用于概率评估.

结论:

训练集的大小和多样性是开发准确和可概括的深度学习模型用于化学转移预测的关键因素.
开发的MPNN方法,特别是使用大型数据集,为光谱分配和验证提供了可靠的工具.
从模型中得出的定量错误估计提高了将观察到的NMR光谱分配给分子结构的信心.

更多相关视频

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Author Spotlight: Advancing Alzheimer's Research – Exploring Early Detection and Multi-Omics Approaches

Published on: December 15, 2023

Pure Shift Nuclear Magnetic Resonance: a New Tool for Plant Metabolomics

Pure Shift Nuclear Magnetic Resonance: a New Tool for Plant Metabolomics

Published on: July 31, 2021

相关概念视频

Carbon-13 (¹³C) NMR: Overview

Carbon-13 (¹³C) NMR: Overview

Carbon-13 is a naturally occurring NMR-active isotope of carbon with a low natural abundance of 1.1%. In contrast, carbon-12 is the most abundant isotope of carbon with zero nuclear spin. Therefore, it is NMR inactive. The gyromagnetic ratio of carbon-13 is smaller than that of protons. As a result, carbon-13 resonance is about 6000 times weaker than proton resonance. For a given magnetic field strength, the resonance frequency of carbon-13 is about one-fourth of the resonance frequency for...

¹³C NMR: ¹H–¹³C Decoupling

¹³C NMR: ¹H–¹³C Decoupling

The probability of having two carbon-13 atoms next to each other is negligible because of the low natural abundance of carbon-13. Consequently, peak splitting due to carbon-carbon spin-spin coupling is not observed in spectra. However, protons up to three sigma bonds away split the carbon signal according to the n+1 rule, resulting in complicated spectra.
A broadband decoupling technique is used to simplify these complex, sometimes overlapping, signals. Broadband decoupling relies on a...

¹H NMR: Interpreting Distorted and Overlapping Signals

¹H NMR: Interpreting Distorted and Overlapping Signals

Spin systems where the difference in chemical shifts of the coupled nuclei is greater than ten times J are called first-order spin systems. These nuclei are weakly coupled, and their chemical shifts and coupling constant can generally be estimated from the well-separated signals in the spectrum.
As Δν decreases and the signals move closer, the doublets appear increasingly distorted. The intensities of the inner lines increase at the cost of those of the outer lines as the signals are...

NMR Spectroscopy: Chemical Shift Overview

NMR Spectroscopy: Chemical Shift Overview

The position of the absorption signal of a sample is reported relative to the position of the signal of tetramethylsilane (TMS), which is added as an internal reference while recording spectra. The difference between the absorption frequencies of the sample and TMS (in Hz) is divided by the spectrometer operating frequency (in MHz) to obtain a dimensionless quantity called the chemical shift. It is reported on the δ (delta) scale and expressed in parts per million.
For instance, the proton...

¹³C NMR: Distortionless Enhancement by Polarization Transfer (DEPT)

¹³C NMR: Distortionless Enhancement by Polarization Transfer (DEPT)

When proton-coupled carbon-13 spectra are simplified by a broadband proton decoupling technique, structural information about the coupled protons is lost. Distortionless enhancement by polarization transfer (DEPT) is a technique that provides information on the number of hydrogens attached to each carbon in a molecule. While the DEPT experiment utilizes complex pulse sequences, the pulse delay and flip angle are specifically manipulated. The resulting signals have different phases depending on...

NMR Spectroscopy Of Amines

NMR Spectroscopy Of Amines

In proton NMR spectroscopy, primary amines and secondary amines showcase their N–H protons as a broad signal in the chemical shift range between δ 0.5 and 5 ppm. The exact position in this range depends on several factors, including sample concentration, hydrogen bonding, and the type of solvent used. Since amine protons undergo fast proton exchange in solution, the protons are labile and therefore do not participate in any splitting with adjacent protons. Thus, the observed peak is...