Pre-processing of three-way data by pulse-coupled neural networks -: an imaging approach

被引：7

作者：

Åberg, KM ^{[1
]}

Jacobsson, SP ^{[1
]}

机构：

[1] Stockholm Univ, Dept Analyt Chem, SE-10691 Stockholm, Sweden

来源：

CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS | 2001年 / 57卷 / 01期

关键词：

QSAR; QSRR; Q-field; partial least squares; neural network; pulse-coupled neural network; steroid;

D O I：

10.1016/S0169-7439(01)00118-6

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

A new method for pre-processing three-dimensional data to model quantitative structure-retention relationships (QSRR) is presented. The pre-processing of three-dimensional images of molecules is done with a pulse-coupled neural network (PCNN). The PCNN is capable of transforming an image to a short time series representation of the molecule, which is more suitable for QSRR modelling with partial least squares than the original data. The method was developed and tested on a steroid data set of 24 compounds with reversed-phase high-performance liquid chromatographic retention data. The QSRR models are stable with respect to the parameters of the PCNN. Test set correlations (q(2)) of 0.95 and cross-validated r(2) of 0.95 are readily obtained. (C) 2001 Elsevier Science B.V. All rights reserved.

引用

页码：25 / 36

页数：12

共 19 条

[1] Test analytes for studies of the molecular mechanism of chromatographic separations by quantitative structure-retention relationships [J].

Al-Haj, MA ;

Kaliszan, R ;

Nasal, A .

ANALYTICAL CHEMISTRY, 1999, 71 (15) :2976-2985

[2] Elimination of uninformative variables for multivariate calibration [J].

Centner, V ;

Massart, DL ;

deNoord, OE ;

deJong, S ;

Vandeginste, BM ;

Sterna, C .

ANALYTICAL CHEMISTRY, 1996, 68 (21) :3851-3858

[3] COMPARATIVE MOLECULAR-FIELD ANALYSIS (COMFA) .1. EFFECT OF SHAPE ON BINDING OF STEROIDS TO CARRIER PROTEINS [J].

CRAMER, RD ;

PATTERSON, DE ;

BUNCE, JD .

JOURNAL OF THE AMERICAN CHEMICAL SOCIETY, 1988, 110 (18) :5959-5967

[4] THE DEVELOPMENT AND USE OF QUANTUM-MECHANICAL MOLECULAR-MODELS .76. AM1 - A NEW GENERAL-PURPOSE QUANTUM-MECHANICAL MOLECULAR-MODEL [J].

DEWAR, MJS ;

ZOEBISCH, EG ;

HEALY, EF ;

STEWART, JJP .

JOURNAL OF THE AMERICAN CHEMICAL SOCIETY, 1985, 107 (13) :3902-3909

[5] Feature Linking via Synchronization among Distributed Assemblies: Simulations of Results from Cat Visual Cortex [J].

Eckhorn, R. ;

Reitboeck, H. J. ;

Arndt, M. ;

Dicke, P. .

NEURAL COMPUTATION, 1990, 2 (03) :293-307

[6] GA strategy for variable selection in QSAR studies: Application of GA-based region selection to a 3D-QSAR study of acetylcholinesterase inhibitors [J].

Hasegawa, K ;

Kimura, T ;

Funatsu, K .

JOURNAL OF CHEMICAL INFORMATION AND COMPUTER SCIENCES, 1999, 39 (01) :112-120

[7] PULSE-COUPLED NEURAL NETS - TRANSLATION, ROTATION, SCALE, DISTORTION, AND INTENSITY SIGNAL INVARIANCE FOR IMAGES [J].

JOHNSON, JL .

APPLIED OPTICS, 1994, 33 (26) :6239-6253

[8] Multidimensional pulse image processing of chemical structure data [J].

Kinser, JM ;

Waldemark, K ;

Lindblad, T ;

Jacobsson, SP .

CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS, 2000, 51 (01) :115-124

[9] MOLECULAR SIMILARITY INDEXES IN A COMPARATIVE-ANALYSIS (COMSIA) OF DRUG MOLECULES TO CORRELATE AND PREDICT THEIR BIOLOGICAL-ACTIVITY [J].

KLEBE, G ;

ABRAHAM, U ;

MIETZNER, T .

JOURNAL OF MEDICINAL CHEMISTRY, 1994, 37 (24) :4130-4146

[10]

Lindeberg T., 1998, IMAGE PROCESSING USI, P30

← 1 2 →