PUBLICATION
MCR-ALS analysis of 1H NMR spectra by segments to study the zebrafish exposure to acrylamide
- Authors
- Pérez, Y., Casado, M., Raldúa, D., Prats, E., Piña, B., Tauler, R., Alfonso, I., Puig-Castellví, F.
- ID
- ZDB-PUB-200704-15
- Date
- 2020
- Source
- Analytical and bioanalytical chemistry 412(23): 5695-5706 (Journal)
- Registered Authors
- Piña, Benjamin, Raldúa, Demetrio
- Keywords
- 1H, Chemometrics, MCR-ALS, Metabolite discovery, Metabolomics, NMR
- MeSH Terms
-
- Zebrafish
- Principal Component Analysis
- Multivariate Analysis
- Acrylamide/toxicity*
- Animals
- Proton Magnetic Resonance Spectroscopy/methods*
- Discriminant Analysis
- Metabolomics
- PubMed
- 32617759 Full text @ Anal. Bioanal. Chem.
Citation
Pérez, Y., Casado, M., Raldúa, D., Prats, E., Piña, B., Tauler, R., Alfonso, I., Puig-Castellví, F. (2020) MCR-ALS analysis of 1H NMR spectra by segments to study the zebrafish exposure to acrylamide. Analytical and bioanalytical chemistry. 412(23):5695-5706.
Abstract
Metabolomics is currently an important field within bioanalytical science and NMR has become a key technique for drawing the full metabolic picture. However, the analysis of 1H NMR spectra of metabolomics samples is often very challenging, as resonances usually overlap in crowded regions, hindering the steps of metabolite profiling and resonance integration. In this context, a pre-processing method for the analysis of 1D 1H NMR data from metabolomics samples is proposed, consisting of the blind resolution and integration of all resonances of the spectral dataset by multivariate curve resolution-alternating least squares (MCR-ALS). The resulting concentration estimates can then be examined with traditional chemometric methods such as principal component analysis (PCA), ANOVA-simultaneous component analysis (ASCA), and partial least squares-discriminant analysis (PLS-DA). Since MCR-ALS does not require the use of spectral templates, the concentration estimates for all resonances are obtained even before being assigned. Consequently, the metabolomics study can be performed without neglecting any relevant resonance. In this work, the proposed pipeline performance was validated with 1D 1H NMR spectra from a metabolomics study of zebrafish upon acrylamide (ACR) exposure. Remarkably, this method represents a framework for the high-throughput analysis of NMR metabolomics data that opens the way for truly untargeted NMR metabolomics analyses. Graphical abstract.
Genes / Markers
Expression
Phenotype
Mutations / Transgenics
Human Disease / Model
Sequence Targeting Reagents
Fish
Orthology
Engineered Foreign Genes
Mapping