PLS‐DA for compositional data with application to metabolomics

Abstract

When quantifying information in metabolomics, the results are often expressed as data carrying only relative information. Vectors of these data have positive components, and the only relevant information is contained in the ratios between their parts; such observations are called compositional data. The aim of the paper is to demonstrate how partial least squares discriminant analysis (PLS‐DA)—the most widely used method in chemometrics for multivariate classification—can be applied to compositional data. Theoretical arguments are provided, and data sets from metabolomics are investigated. The data are related to the diagnosis of inherited metabolic disorders (IMDs). The first example analyzes the significance of the corresponding regression parameters (metabolites) using a small data set resulting from targeted metabolomics, where just a subset of potential markers is selected. The second example—the approach of untargeted metabolomics—was used for the analysis detecting almost 500 metabolites. The significance of the metabolites is investigated by applying PLS‐DA, accommodated according to a compositional approach. The significance of important metabolites (markers of diseases) is more clearly visible with the compositional method in both examples. Also, cross‐validation methods lead to better results in case of using the compositional approach. Copyright © 2014 John Wiley & Sons, Ltd.

Normalization techniques for PARAFAC modeling of urine metabolomic data

“…In each row the score plot (a, d) and two loading plots (b, c, e, f) are displayed ''Introduction'' section) into account. According to recent experiences, the information obtained using clr coordinates could be even further enhanced, e.g., by taking orthonormal coordinates (Egozcue et al 2003;Kalivodová et al 2015) or by considering measurement errors in metabolites. The latter case results in weighting of compositional parts as generalization of clr variables in order to reduce possible false positives in metabolomical data sets.…”

“…The aim of this paper is to develop PARAFAC models in the context of compositional data analysis for statistical processing of urine metabolomics data, that was recently also successfully applied in other metabolomic contexts (Janečková et al 2012;Kalivodová et al 2015;Hron et al 2012;Korhoňová et al 2009) and to discuss also alternative normalization and transformation techniques. In the next sections, we will explain the theory and will illustrate that with a practical example of urine metabolomic data of newborn children suffering from asphyxia, measured at five time points, and two small simulation studies.…”

Macranthoidin B Modulates Key Metabolic Pathways to Enhance ROS Generation and Induce Cytotoxicity and Apoptosis in Colorectal Cancer

“…PLS-DA is a powerful method for compositional (metabolic) data that contains more metabolites (in hundreds) than biological materials (only tens) [29]. The results showed that levels of 31 metabolites (Table 1; Fig.…”

Practical Aspects of Log-ratio Coordinate Representations in Regression with Compositional Response

“…The first one concerns special orthonormal coordinate systems that enable interpretation in terms of the original compositional parts (with respect to the other parts in the actual composition) and were applied in a number of applications including regression modelling [10,12,14]. Although it is theoretically sound to work exclusively in orthonormal coordinates, this particular choice DOI: 10.1515/msr-2016-0029 of coordinates seems to be also a bit impractical as for a Dpart composition D coordinate systems are needed.…”