Abstract:Metabolomics is a dynamic and emerging research field, similar to proteomics, transcriptomics and genomics in affording global understanding of biological systems. It is particularly useful in functional genomic studies in which metabolism is thought to be perturbed. Metabolomics provides a snapshot of the metabolic dynamics that reflect the response of living systems to both pathophysiological stimuli and/or genetic modification. Because this approach makes possible the examination of interactions between an organism and its diet or environment, it is particularly useful for identifying biomarkers of disease processes that involve the environment. For example, the interaction of a high fat diet with cardiovascular disease can be studied via such a metabolomics approach by modeling the interaction between genes and diet. The high reproducibility of NMR-based techniques gives this method a number of advantages over other analytical techniques in large-scale and long-term metabolomic studies, such as epidemiological studies. This approach has been used to study a wide range of diseases, through the examination of biofluids, including blood plasma/serum, urine, blister fluid, saliva and semen, as well as tissue extracts and intact tissue biopsies. However, complicating the use of NMR spectroscopy in biomarker discovery is the fact that numerous variables can effect metabolic composition including, fasting, stress, drug administration, diet, gender, age, physical activity, life style and the subject’s health condition. To minimize the influence of these variations in the datasets, all experimental conditions including sample collection, storage, preparation as well as NMR spectroscopic parameters and data analysis should be optimized carefully and conducted in an identical manner as described by the local standard operating protocol . This review highlights the potential applications of NMR-based metabolomics studies and gives some recommendations to improve sample collection, sample preparation and data analysis in using this approach.

The development of machine learning approaches in two-dimensional NMR data interpretation for metabolomics applications

Applications of machine learning in metabolomics: Disease modeling and classification

Multivariate analysis of NMR‐based metabolomic data

Integrating Machine Learning in Metabolomics: A Path to Enhanced Diagnostics and Data Interpretation

Deep Neural Networks for Classification of LC-MS Spectral Peaks.

Automated metabolic assignment: Semi-supervised learning in metabolic analysis employing two dimensional Nuclear Magnetic Resonance (NMR)

MDAS: An integrated system for metabonomic data analysis

An Efficient Approach for Obtaining Small and Macro-molecular 1H NMR Spectra Based on Neural Network

Toward Diagnosis Of Diabetes By Nmr And Multivariate Analysis: Study Of 1h Nmr Spectra Of Human Serum In An Integrated Working Environment

Machine Learning Using Neural Networks for Metabolomic Pathway Analyses

The application of artificial neural networks in metabolomics: a historical perspective

Non-negative Principal Component Analysis for NMR-based Metabolomic Data Analysis

Improved Prediction of Carbonless NMR Spectra by the Machine Learning of Theoretical and Fragment Descriptors for Environmental Mixture Analysis

NMR Spectroscopy for Metabolomics Research

Application of sparse linear discriminant analysis for metabolomics data

Enhanced Detection And Annotation Of Small Molecules In Metabolomics Using Molecular-Network-Oriented Parameter Optimization

A New Data Processing Method for Metabonomic and Its Application in a Study of Diabetes

Deep metabolome: Applications of deep learning in metabolomics

Accurate, fully-automated NMR spectral profiling for metabolomics

Using neural networks to obtain NMR spectra of both small and macromolecules from blood samples in a single experiment

NMR-based metabolomics in human disease diagnosis: applications, limitations, and recommendations