CROSS-PEAK CLASSIFICATION IN 2-DIMENSIONAL NUCLEAR-MAGNETIC-RESONANCE SPECTRA USING A 2-LAYER NEURAL-NETWORK

被引：9

作者：

CORNE, SA

FISHER, J

JOHNSON, AP

NEWELL, WR

机构：

[1] School of Chemistry, University of Leeds, Leeds

来源：

ANALYTICA CHIMICA ACTA | 1993年 / 278卷 / 01期

关键词：

NUCLEAR MAGNETIC RESONANCE SPECTROMETRY; CROSSPEAK CLASSIFICATION; NEURAL NETWORKS; 2-LAYER NEURAL NETWORK;

D O I：

10.1016/0003-2670(93)80093-Z

中图分类号：

O65 [分析化学];

学科分类号：

070302 ; 081704 ;

摘要：

A two-layer simulated neural network has been trained to classify cross-peaks in two-dimensional nuclear magnetic resonance spectra. Examples of peaks for both network training and testing were selected by an experienced spectroscopist. The trained network has been used to classify previously unseen data. Spectral artefacts and authentic cross-peaks are distinguished. Peaks whose shapes have been modified, for example by overlap, are classified correctly. A spectrum for phoratoxin B, a protein of 46 amino acid residues, is used to illustrate the training and performance of the network.

引用

页码：149 / 158

页数：10

共 18 条

[1]

Beale R, 1990, NEURAL COMPUTING INT

[2] COMPARISON OF THE TRAINING OF NEURAL NETWORKS FOR QUANTITATIVE X-RAY-FLUORESCENCE SPECTROMETRY BY A GENETIC ALGORITHM AND BACKWARD ERROR PROPAGATION [J].