Related Experiment Video
Updated: May 22, 2026
11:02
Identification and Quantification of Deranged Metabolites in Critically Ill Patients Using NMR-Based Metabolomics
Published on: November 29, 2024
State-of-the art data normalization methods improve NMR-based metabolomic analysis
Stefanie M Kohl1, Matthias S Klein, Jochen Hochrein
1Institute of Functional Genomics, University of Regensburg, Josef-Engert-Strasse 9, 93053 Regensburg, Germany.
Summary
Data preprocessing is crucial for metabolomic analysis. Quantile and Cubic-Spline Normalization methods, originally for DNA microarrays, best reduce bias and improve sample classification in NMR spectroscopy data.
Area of Science:
- Metabolomics
- Biomedical data analysis
- Nuclear Magnetic Resonance (NMR) spectroscopy
Background:
- Metabolomic data analysis presents challenges in identifying differential metabolites, estimating fold changes, and classifying samples.
- Effective data preprocessing is essential to minimize bias and experimental variance before downstream analysis.
- Normalization methods aim to reduce sample-to-sample variation and adjust metabolite variances.
Purpose of the Study:
- To systematically compare different data normalization methods for metabolomic datasets.
- To evaluate the impact of normalization on sample classification and differential metabolite screening.
- To identify optimal preprocessing strategies for NMR-based metabolomic studies.
Main Methods:
- Comparison of two types of normalization methods: sample-to-sample variation removal and variable scaling/variance stabilization.
- Application of methods to two distinct datasets: urinary NMR fingerprints from healthy and Autosomal Polycystic Kidney Disease (ADPKD) patients, and a spiked urine matrix.
- Evaluation of preprocessing impact on sample classification, differential metabolite screening, and data structure.
Main Results:
- Quantile and Cubic-Spline Normalization, adapted from DNA microarray analysis, demonstrated superior performance.
- These methods effectively reduced bias, improved the accuracy of fold change detection, and enhanced sample classification.
- Preprocessing significantly influenced the structure of the analyzed metabolomic data.
Conclusions:
- Preprocessing methods originally developed for DNA microarray analysis, specifically Quantile and Cubic-Spline Normalization, are highly effective for NMR-based metabolomic data.
- These methods offer significant improvements in reducing bias, detecting differential metabolites, and classifying samples.
- Optimal data preprocessing is critical for reliable and accurate interpretation of complex metabolomic datasets.
