Bylesjö Max
Fios Genomics Ltd, Nine Edinburgh Bioquarter, 9 Little France Road, Edinburgh, EH16 4UX, UK,
Methods Mol Biol. 2015;1277:137-46. doi: 10.1007/978-1-4939-2377-9_11.
Metabonomics aims to identify and quantify all small-molecule metabolites in biologically relevant samples using high-throughput techniques such as NMR and chromatography/mass spectrometry. This generates high-dimensional data sets with properties that require specialized approaches to data analysis. This chapter describes multivariate statistics and analysis tools to extract meaningful information from metabonomic data sets. The focus is on the use and interpretation of latent variable methods such as principal component analysis (PCA), partial least squares/projections to latent structures (PLS), and orthogonal PLS (OPLS). Descriptions of the key steps of the multivariate data analyses are provided with demonstrations from example data.
代谢组学旨在利用核磁共振（NMR）和色谱/质谱等高通量技术，识别和定量生物相关样品中的所有小分子代谢物。这会生成具有特定属性的高维数据集，需要采用专门的数据分析方法。本章介绍多元统计和分析工具，以从代谢组学数据集中提取有意义的信息。重点是潜在变量方法的使用和解释，如主成分分析（PCA）、偏最小二乘法/潜在结构投影（PLS）和正交PLS（OPLS）。通过示例数据演示，提供了多元数据分析关键步骤的描述。
