Linear-mixed effects models for feature selection in high-dimensional NMR spectra

Yajun Mei, Seoung Bum Kim, Kwok Leung Tsui

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

Feature selection in metabolomics can identify important metabolite features that play a significant role in discriminating between various conditions among samples. In this paper, we propose an efficient feature selection method for high-resolution nuclear magnetic resonance (NMR) spectra obtained from time-course experiments. Our proposed approach combines linear-mixed effects (LME) models with a multiple testing procedure based on a false discovery rate. The proposed LME approach is illustrated using NMR spectra with 574 metabolite features obtained for an experiment to examine metabolic changes in response to sulfur amino acid intake. The experimental results showed that classification models constructed with the features selected by the proposed approach resulted in lower rates of misclassification than those models with full features. Furthermore, we compared the LME approach with the two-sample t-test approach that oversimplifies the time-course factor.

Original languageEnglish
Pages (from-to)4703-4708
Number of pages6
JournalExpert Systems with Applications
Volume36
Issue number3 PART 1
DOIs
Publication statusPublished - 2009 Apr 1
Externally publishedYes

Fingerprint

Feature extraction
Nuclear magnetic resonance
Metabolites
Amino acids
Sulfur
Experiments
Testing
Metabolomics

Keywords

  • False discovery rate
  • Feature selection
  • Linear-mixed effects models
  • Multiple hypothesis testing
  • Nuclear magnetic resonance

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Engineering(all)

Cite this

Linear-mixed effects models for feature selection in high-dimensional NMR spectra. / Mei, Yajun; Kim, Seoung Bum; Tsui, Kwok Leung.

In: Expert Systems with Applications, Vol. 36, No. 3 PART 1, 01.04.2009, p. 4703-4708.

Research output: Contribution to journalArticle

@article{b35e6159a42f4d92821198271af10aeb,
title = "Linear-mixed effects models for feature selection in high-dimensional NMR spectra",
abstract = "Feature selection in metabolomics can identify important metabolite features that play a significant role in discriminating between various conditions among samples. In this paper, we propose an efficient feature selection method for high-resolution nuclear magnetic resonance (NMR) spectra obtained from time-course experiments. Our proposed approach combines linear-mixed effects (LME) models with a multiple testing procedure based on a false discovery rate. The proposed LME approach is illustrated using NMR spectra with 574 metabolite features obtained for an experiment to examine metabolic changes in response to sulfur amino acid intake. The experimental results showed that classification models constructed with the features selected by the proposed approach resulted in lower rates of misclassification than those models with full features. Furthermore, we compared the LME approach with the two-sample t-test approach that oversimplifies the time-course factor.",
keywords = "False discovery rate, Feature selection, Linear-mixed effects models, Multiple hypothesis testing, Nuclear magnetic resonance",
author = "Yajun Mei and Kim, {Seoung Bum} and Tsui, {Kwok Leung}",
year = "2009",
month = "4",
day = "1",
doi = "10.1016/j.eswa.2008.06.032",
language = "English",
volume = "36",
pages = "4703--4708",
journal = "Expert Systems with Applications",
issn = "0957-4174",
publisher = "Elsevier Limited",
number = "3 PART 1",

}

TY - JOUR

T1 - Linear-mixed effects models for feature selection in high-dimensional NMR spectra

AU - Mei, Yajun

AU - Kim, Seoung Bum

AU - Tsui, Kwok Leung

PY - 2009/4/1

Y1 - 2009/4/1

N2 - Feature selection in metabolomics can identify important metabolite features that play a significant role in discriminating between various conditions among samples. In this paper, we propose an efficient feature selection method for high-resolution nuclear magnetic resonance (NMR) spectra obtained from time-course experiments. Our proposed approach combines linear-mixed effects (LME) models with a multiple testing procedure based on a false discovery rate. The proposed LME approach is illustrated using NMR spectra with 574 metabolite features obtained for an experiment to examine metabolic changes in response to sulfur amino acid intake. The experimental results showed that classification models constructed with the features selected by the proposed approach resulted in lower rates of misclassification than those models with full features. Furthermore, we compared the LME approach with the two-sample t-test approach that oversimplifies the time-course factor.

AB - Feature selection in metabolomics can identify important metabolite features that play a significant role in discriminating between various conditions among samples. In this paper, we propose an efficient feature selection method for high-resolution nuclear magnetic resonance (NMR) spectra obtained from time-course experiments. Our proposed approach combines linear-mixed effects (LME) models with a multiple testing procedure based on a false discovery rate. The proposed LME approach is illustrated using NMR spectra with 574 metabolite features obtained for an experiment to examine metabolic changes in response to sulfur amino acid intake. The experimental results showed that classification models constructed with the features selected by the proposed approach resulted in lower rates of misclassification than those models with full features. Furthermore, we compared the LME approach with the two-sample t-test approach that oversimplifies the time-course factor.

KW - False discovery rate

KW - Feature selection

KW - Linear-mixed effects models

KW - Multiple hypothesis testing

KW - Nuclear magnetic resonance

UR - http://www.scopus.com/inward/record.url?scp=58349115402&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=58349115402&partnerID=8YFLogxK

U2 - 10.1016/j.eswa.2008.06.032

DO - 10.1016/j.eswa.2008.06.032

M3 - Article

VL - 36

SP - 4703

EP - 4708

JO - Expert Systems with Applications

JF - Expert Systems with Applications

SN - 0957-4174

IS - 3 PART 1

ER -