Robust biomarker identification in a two-class problem based on pairwise log-ratios

Walach, Jan; Filzmoser, Peter; Hron, Karel; Walczak, Beata; Najdekr, Lukáš

doi:10.1016/j.chemolab.2017.09.003

Cited by 20 publications

(21 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…After transformation, the analysis then proceeds as if the data were absolute, but with a caveat: the interpretation of the results depends on the reference used. Second, the "pragmatic approach" analyzes pairwise log ratios directly; this type of analysis has been used to score important genes (15) and gene pairs (16,17), and to reduce the dimensionality of the data (17). This approach makes sense when the ratios themselves have some importance to the analyst.…”

mentioning

confidence: 99%

Interpretable Log Contrasts for the Classification of Health Biomarkers: a New Approach to Balance Selection

Quinn¹,

Erb

2020

mSystems

View full text Add to dashboard Cite

Since the turn of the century, technological advances have made it possible to obtain the molecular profile of any tissue in a cost-effective manner. Among these advances are sophisticated high-throughput assays that measure the relative abundances of microorganisms, RNA molecules, and metabolites. While these data are most often collected to gain new insights into biological systems, they can also be used as biomarkers to create clinically useful diagnostic classifiers. How best to classify high-dimensional -omics data remains an area of active research. However, few explicitly model the relative nature of these data and instead rely on cumbersome normalizations. This report (i) emphasizes the relative nature of health biomarkers, (ii) discusses the literature surrounding the classification of relative data, and (iii) benchmarks how different transformations perform for regularized logistic regression across multiple biomarker types. We show how an interpretable set of log contrasts, called balances, can prepare data for classification. We propose a simple procedure, called discriminative balance analysis, to select groups of 2 and 3 bacteria that can together discriminate between experimental conditions. Discriminative balance analysis is a fast, accurate, and interpretable alternative to data normalization. IMPORTANCE High-throughput sequencing provides an easy and cost-effective way to measure the relative abundance of bacteria in any environmental or biological sample. When these samples come from humans, the microbiome signatures can act as biomarkers for disease prediction. However, because bacterial abundance is measured as a composition, the data have unique properties that make conventional analyses inappropriate. To overcome this, analysts often use cumbersome normalizations. This article proposes an alternative method that identifies pairs and trios of bacteria whose stoichiometric presence can differentiate between diseased and nondiseased samples. By using interpretable log contrasts called balances, we developed an entirely normalization-free classification procedure that reduces the feature space and improves the interpretability, without sacrificing classifier performance.

show abstract

mentioning

confidence: 99%

Interpretable Log Contrasts for the Classification of Health Biomarkers: a New Approach to Balance Selection

Quinn¹,

Erb

2020

mSystems

View full text Add to dashboard Cite

show abstract

“…The method rPLR has been developed for the identification of biomarkers using binary classification . The method rPLR is independent from the absolute values of the M variables because it uses the log ratios of all pairs ( j , k ) of the variables; for object i given by ln( x ij / x ik ) with j , k = 1, …, M .…”

Section: Data Analysis Methodsmentioning

confidence: 99%

“…A variable j with high importance for classification (a biomarker) will have considerably different elements 1 t jk and 2 t jk and tentatively smaller than t jk in T (for all k ). A statistic V j for estimation of the importance of variable j has been proposed as

V_{j} = sum ({}, ([], N_{1 1} {t_{italicjk}}^{0.5} + N_{2 2} {t_{italicjk}}^{0.5}) / {italicNt}_{italicjk}^{0.5}) sum for k = 1, \dots, M .

V j is approximately normally distributed, as well as the normalized version suggested for practical use:

V_{j}^{*} = - ([], V_{j} - italicmean ((), V_{k})) / italicsd ((), V_{k}),

with mean for the arithmetic mean, sd for the standard deviation, and k = 1, …, M . Big values for V * indicate a high variable importance (owing to the minus sign in the above equation); a reasonable cutoff is the 0.975 quantile.…”

Section: Data Analysis Methodsmentioning

confidence: 99%

Significance of variables for discrimination: Applied to the search of organic ions in mass spectra measured on cometary particles

Вармуза

Filzmoser

Hoffmann

et al. 2018

Journal of Chemometrics

Self Cite

View full text Add to dashboard Cite

The instrument Cometary Secondary Ion Mass Analyzer (COSIMA) on board of the European Space Agency mission Rosetta to the comet 67P/Churyumov-Gerasimenko is a secondary ion mass spectrometer with a time-of-flight mass analyzer. It collected near the comet several thousand particles, imaged them, and analyzed the elemental and chemical compositions of their surfaces. In this study, variables have been generated from the spectral data covering the mass ranges of potential C-, H-, N-, and O-containing ions. The variable importance in binary discriminations between spectra measured on cometary particles and those measured on the target background has been estimated by the univariate t test and the multivariate methods discriminant partial least squares, random forest, and a robust method based on the log ratios of all variable pairs. The results confirm the presence of organic substances in cometary matter-probably a complex macromolecular mixture. KEYWORDS classification, comet 67P/Churyumov-Gerasimenko, D-PLS, pairwise log ratios, random forest, Rosetta, time-of-flight secondary ion mass spectrometry, variable importance

show abstract

“…The methods above depend on a log-ratio transformation to standardize the comparison of one gene's expression (or one pair's coordination) with another. However, by comparing the variance of the log-ratios (VLR) within groups to the total VLR, we do not need a reference to estimate between-group differences in coordination [15,66]:…”

Section: Part 2b: Transformation-independent Analysesmentioning

confidence: 99%

A field guide for the compositional analysis of any-omics data

Quinn

Erb

Gloor

et al. 2018

Preprint

View full text Add to dashboard Cite

Next-generation sequencing (NGS) has made it possible to determine the sequence and relative abundance of all nucleotides in a biological or environmental sample. Today, NGS is routinely used to understand many important topics in biology from human disease to microorganism diversity. A cornerstone of NGS is the quantification of RNA or DNA presence as counts. However, these counts are not counts per se: the magnitude of the counts are determined arbitrarily by the sequencing depth, not by the input material. Consequently, counts must undergo normalization prior to use. Conventional normalization methods require a set of assumptions: they assume that the majority of features are unchanged, and that all environments under study have the same carrying capacity for nucleotide synthesis. These assumptions are often untestable and may not hold when comparing heterogeneous samples (e.g., samples collected across distinct cancers or tissues). Instead, methods developed within the field of compositional data analysis offer a general solution that is assumption-free and valid for all data. In this manuscript, we synthesize the extant literature to provide a concise guide on how to apply compositional data analysis to NGS count data. In doing so, we review zero replacement, differential abundance analysis, and within-group and between-group coordination analysis. We then discuss how this pipeline can accommodate complex study design, facilitate the analysis of vertically and horizontally integrated data, including multiomics data, and further extend to single-cell sequencing data. In highlighting the limitations of total library size, effective library size, and spike-in normalizations, we propose the log-ratio transformation as a general solution to answer the question, "Relative to some important activity of the cell, what is changing?". Taken together, this manuscript establishes the first fully comprehensive analysis protocol that is suitable for any and all -omics data.

show abstract

Robust biomarker identification in a two-class problem based on pairwise log-ratios

Cited by 20 publications

References 22 publications

Interpretable Log Contrasts for the Classification of Health Biomarkers: a New Approach to Balance Selection

Interpretable Log Contrasts for the Classification of Health Biomarkers: a New Approach to Balance Selection

Significance of variables for discrimination: Applied to the search of organic ions in mass spectra measured on cometary particles

A field guide for the compositional analysis of any-omics data

Contact Info

Product

Resources

About