ANOVA-Like Differential Expression (ALDEx) Analysis for Mixed Population RNA-Seq

Abstract

Experimental variance is a major challenge when dealing with high-throughput sequencing data. This variance has several sources: sampling replication, technical replication, variability within biological conditions, and variability between biological conditions. The high per-sample cost of RNA-Seq often precludes the large number of experiments needed to partition observed variance into these categories as per standard ANOVA models. We show that the partitioning of within-condition to between-condition variation cannot reasonably be ignored, whether in single-organism RNA-Seq or in Meta-RNA-Seq experiments, and further find that commonly-used RNA-Seq analysis tools, as described in the literature, do not enforce the constraint that the sum of relative expression levels must be one, and thus report expression levels that are systematically distorted. These two factors lead to misleading inferences if not properly accommodated. As it is usually only the biological between-condition and within-condition differences that are of interest, we developed ALDEx, an ANOVA-like differential expression procedure, to identify genes with greater between- to within-condition differences. We show that the presence of differential expression and the magnitude of these comparative differences can be reasonably estimated with even very small sample sizes.

Metabolic derangements identified through untargeted metabolomics in a cross-sectional study of Nigerian children with severe acute malnutrition

“…To avoid inappropriate statistical inferences made from compositional data, centred log-ratios (clr), a method previously described by Aitchison (Aitchison 1982), and adapted to microbiome data was used with unpaired Wilcoxon tests for comparisons of OTU level data (Fernandes et al 2013;Fernandes et al 2014). The Benjamini Hochberg (FDR) method was used to control for multiple testing with a significance threshold of 0.1.…”

Compositional analysis: a valid approach to analyze microbiome high-throughput sequencing data

“…Thus, it is important to ensure that any analysis takes this random component into account (Fernandes et al 2013).…”

“…Compositional data are a term used to describe a data set in which the parts in each sample have an arbitrary or noninformative sum (Aitchison 1986), such as data obtained from high-throughput DNA sequencing (Friedman and Alm 2012;Fernandes et al 2013Fernandes et al , 2014. These data have long been known to be problematic (Pearson 1896), and we now understand that multivariate data analysis approaches such as ordination and clustering and univariate methods that measure differential abundance are invalid (Aitchison 1986;Warton et al 2012;Friedman and Alm 2012;Fernandes et al 2013;.…”

“…These data have long been known to be problematic (Pearson 1896), and we now understand that multivariate data analysis approaches such as ordination and clustering and univariate methods that measure differential abundance are invalid (Aitchison 1986;Warton et al 2012;Friedman and Alm 2012;Fernandes et al 2013;.…”

“…correlation networks, principal component analysis, and others (Pearson 1896;Aitchison 1986). Note that this distortion will also lead to false univariate inferences as well (Fernandes et al 2013(Fernandes et al , 2014.…”