Peak-finding partial least squares for the analysis of 1H NMR spectra

L. P. Ammann, M. Merritt, Arthur I Sagalowsky, P. Nurenberg

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Metabonomic analysis of biofluids and extracts of biological tissues is increasingly being used to diagnose important metabolic differences induced by toxicity, disease processes or genetic differences. 1H nuclear magnetic resonance (NMR) has been shown to be very useful for monitoring the low-molecular weight metabolite milieu typical of many systems. In this paper, a rigorous comparison of five different methods of data reduction and classification has been made. The five methods include principal components analysis (PCA) followed by linear discriminant analysis (LDA), PCA followed by logistic regression, a combined peak-picking-PCA and LDA algorithm, partial least squares (PLS), and a peak-picking PLS algorithm. To evaluate these five methods, a data set consisting of 1H NMR spectra of the extracts of 29 malignant renal tumors and 17 normal tissues were analyzed. It was determined that peak-picking with PLS was the most efficient algorithm for correctly classifying this data set. Also, the peak-picking algorithm makes identification of the metabolites responsible for establishing class membership easier than with the other methods. A variety of different metabolites, including several amino acids and choline containing compounds were identified as markers for malignancy.

Original languageEnglish (US)
Pages (from-to)231-238
Number of pages8
JournalJournal of Chemometrics
Volume20
Issue number6-7
DOIs
StatePublished - Jun 2006

Fingerprint

Nuclear Magnetic Resonance
Partial Least Squares
Metabolites
Nuclear magnetic resonance
Principal component analysis
Principal Component Analysis
Discriminant analysis
Discriminant Analysis
Tissue
Biological Tissue
Data Classification
Data Reduction
Least Square Algorithm
Logistic Regression
Toxicity
Choline
Amino Acids
Logistics
Amino acids
Tumors

Keywords

  • Linear discriminant analysis
  • Metabonomics
  • NMR
  • Partial least squares
  • Peak-picking

ASJC Scopus subject areas

  • Analytical Chemistry
  • Statistics and Probability

Cite this

Peak-finding partial least squares for the analysis of 1H NMR spectra. / Ammann, L. P.; Merritt, M.; Sagalowsky, Arthur I; Nurenberg, P.

In: Journal of Chemometrics, Vol. 20, No. 6-7, 06.2006, p. 231-238.

Research output: Contribution to journalArticle

Ammann, L. P. ; Merritt, M. ; Sagalowsky, Arthur I ; Nurenberg, P. / Peak-finding partial least squares for the analysis of 1H NMR spectra. In: Journal of Chemometrics. 2006 ; Vol. 20, No. 6-7. pp. 231-238.
@article{a5428583dce9449fae3293bce42ce3fa,
title = "Peak-finding partial least squares for the analysis of 1H NMR spectra",
abstract = "Metabonomic analysis of biofluids and extracts of biological tissues is increasingly being used to diagnose important metabolic differences induced by toxicity, disease processes or genetic differences. 1H nuclear magnetic resonance (NMR) has been shown to be very useful for monitoring the low-molecular weight metabolite milieu typical of many systems. In this paper, a rigorous comparison of five different methods of data reduction and classification has been made. The five methods include principal components analysis (PCA) followed by linear discriminant analysis (LDA), PCA followed by logistic regression, a combined peak-picking-PCA and LDA algorithm, partial least squares (PLS), and a peak-picking PLS algorithm. To evaluate these five methods, a data set consisting of 1H NMR spectra of the extracts of 29 malignant renal tumors and 17 normal tissues were analyzed. It was determined that peak-picking with PLS was the most efficient algorithm for correctly classifying this data set. Also, the peak-picking algorithm makes identification of the metabolites responsible for establishing class membership easier than with the other methods. A variety of different metabolites, including several amino acids and choline containing compounds were identified as markers for malignancy.",
keywords = "Linear discriminant analysis, Metabonomics, NMR, Partial least squares, Peak-picking",
author = "Ammann, {L. P.} and M. Merritt and Sagalowsky, {Arthur I} and P. Nurenberg",
year = "2006",
month = "6",
doi = "10.1002/cem.977",
language = "English (US)",
volume = "20",
pages = "231--238",
journal = "Journal of Chemometrics",
issn = "0886-9383",
publisher = "John Wiley and Sons Ltd",
number = "6-7",

}

TY - JOUR

T1 - Peak-finding partial least squares for the analysis of 1H NMR spectra

AU - Ammann, L. P.

AU - Merritt, M.

AU - Sagalowsky, Arthur I

AU - Nurenberg, P.

PY - 2006/6

Y1 - 2006/6

N2 - Metabonomic analysis of biofluids and extracts of biological tissues is increasingly being used to diagnose important metabolic differences induced by toxicity, disease processes or genetic differences. 1H nuclear magnetic resonance (NMR) has been shown to be very useful for monitoring the low-molecular weight metabolite milieu typical of many systems. In this paper, a rigorous comparison of five different methods of data reduction and classification has been made. The five methods include principal components analysis (PCA) followed by linear discriminant analysis (LDA), PCA followed by logistic regression, a combined peak-picking-PCA and LDA algorithm, partial least squares (PLS), and a peak-picking PLS algorithm. To evaluate these five methods, a data set consisting of 1H NMR spectra of the extracts of 29 malignant renal tumors and 17 normal tissues were analyzed. It was determined that peak-picking with PLS was the most efficient algorithm for correctly classifying this data set. Also, the peak-picking algorithm makes identification of the metabolites responsible for establishing class membership easier than with the other methods. A variety of different metabolites, including several amino acids and choline containing compounds were identified as markers for malignancy.

AB - Metabonomic analysis of biofluids and extracts of biological tissues is increasingly being used to diagnose important metabolic differences induced by toxicity, disease processes or genetic differences. 1H nuclear magnetic resonance (NMR) has been shown to be very useful for monitoring the low-molecular weight metabolite milieu typical of many systems. In this paper, a rigorous comparison of five different methods of data reduction and classification has been made. The five methods include principal components analysis (PCA) followed by linear discriminant analysis (LDA), PCA followed by logistic regression, a combined peak-picking-PCA and LDA algorithm, partial least squares (PLS), and a peak-picking PLS algorithm. To evaluate these five methods, a data set consisting of 1H NMR spectra of the extracts of 29 malignant renal tumors and 17 normal tissues were analyzed. It was determined that peak-picking with PLS was the most efficient algorithm for correctly classifying this data set. Also, the peak-picking algorithm makes identification of the metabolites responsible for establishing class membership easier than with the other methods. A variety of different metabolites, including several amino acids and choline containing compounds were identified as markers for malignancy.

KW - Linear discriminant analysis

KW - Metabonomics

KW - NMR

KW - Partial least squares

KW - Peak-picking

UR - http://www.scopus.com/inward/record.url?scp=33947433788&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33947433788&partnerID=8YFLogxK

U2 - 10.1002/cem.977

DO - 10.1002/cem.977

M3 - Article

AN - SCOPUS:33947433788

VL - 20

SP - 231

EP - 238

JO - Journal of Chemometrics

JF - Journal of Chemometrics

SN - 0886-9383

IS - 6-7

ER -