Chemometric applications in metabolomic studies using chromatography-mass spectrometry

被引：54

作者：

Paul, Alessandra ^{[1
]}

Harrington, Peter de Boves ^{[1
]}

机构：

[1] Ohio Univ, Dept Chem & Biochem, Clippinger Labs, 100 Univ Terrace, Athens, OH 45701 USA

来源：

TRAC-TRENDS IN ANALYTICAL CHEMISTRY | 2021年 / 135卷

基金：

美国国家科学基金会; 加拿大自然科学与工程研究理事会;

关键词：

Chemometrics; Mass spectrometry; Metabolomics; Metabonomics; Data analysis; Bioinformatics; Statistics; Machine learning; Chemical profiling;

D O I：

10.1016/j.trac.2020.116165

中图分类号：

O65 [分析化学];

学科分类号：

070302 ; 081704 ;

摘要：

Metabolomic studies generate large and exceptionally complex datasets. The chemical diversity that exists within the metabolome presents an immense analytical challenge. Chemometric methods are frequently used to analyze such data because these approaches offer an efficient route to meaningful interpretation of results. The techniques used in recent research fall into three general categories: statistical methods, machine learning models, and custom solutions. There are drawbacks and strengths to every approach, and the right choice varies study to study, depending on the experimental design and hypothesis. It is common for researchers to employ multiple methods by building a pipeline of data analysis steps to analyze their spectra. These pipelines are designed to parse the data in a systematic fashion to best answer the question at hand. This review covers advancements in chemometric techniques applied to metabolomics studies in the last five years. (C) 2021 Elsevier B.V. All rights reserved.

引用

页数：9

共 58 条

[1]

Ahrens H., 1971, LINEAR MODELS, V16, P78, DOI [10.1002/bimj.19740160113, DOI 10.1002/BIMJ.19740160113]

[2] Integrating bioinformatics approaches for a comprehensive interpretation of metabolomics datasets [J].