Pairwise comparisons across species are problematic when analyzing functional genomic data

Abstract

There is considerable interest in comparing functional genomic data across species. One goal of such work is to provide an integrated understanding of genome and phenotype evolution. Most comparative functional genomic studies have relied on multiple pairwise comparisons between species, an approach that does not incorporate information about the evolutionary relationships among species. The statistical problems that arise from not considering these relationships can lead pairwise approaches to the wrong conclusions, and are a missed opportunity to learn about biology that can only be understood in an explicit phylogenetic context. Here we examine two recently published studies that compare gene expression across species with pairwise methods, and find reason to question the original conclusions of both. One study interpreted pairwise comparisons of gene expression as support for the ortholog conjecture, the hypothesis that orthologs tend to be more similar than paralogs. The other study interpreted pairwise comparisons of embryonic gene expression across distantly related animals as evidence for a distinct evolutionary process that gave rise to phyla. In each study, distinct patterns of pairwise similarity among species were originally interpreted as evidence of particular evolutionary processes, but instead we find they reflect species relationships. These reanalyses concretely demonstrate the inadequacy of pairwise comparisons for analyzing functional genomic data across species. It will be critical to adopt phylogenetic comparative methods in future functional genomic work. Fortunately, phylogenetic comparative biology is also a rapidly advancing field with many methods that can be directly applied to functional genomic data. SignificanceComparisons of genome function between species are providing important insight into the evolutionary origins of diversity. Here we demonstrate that comparative functional genomics studies can come to the wrong conclusions if they do not take the relationships of species into account and instead rely on pairwise comparisons between species, as is common practice. We re-examined two previously published studies and found problems with pairwise comparisons that draw both their original conclusions into question. One study had found support for the ortholog conjecture and the other had concluded that the evolution of gene expression was different between animal phyla than within them. Our results demonstrate that to answer evolutionary questions about genome function, it is critical to consider evolutionary relationships.

Resolving the backbone of the Brassicaceae phylogeny for investigating trait diversity

“…4a, Table S10). Comparing gene expression across species requires prior knowledge of phylogenetic relationships (Dunn et al, 2017) to account for neutral fluctuations in gene expression levels in different lineages (Rohlfs et al, 2014). Such analyses could not previously be performed with sufficient confidence because (1) branch lengths, as required for downstream analyses, could not be estimated accurately, and (2) the relationship among taxa with newly generated transcriptomes, in particular the placement of the emerging model species Arabis alpina, has not been resolved consistently before our phylogenomic study (Willing et al, 2015).…”

The Ortholog Conjecture Revisited: the Value of Orthologs and Paralogs in Function Prediction

“…As a decrease in structural [41,37] and functional [23,35,34,4,9,32] similarity with divergence is a widely expected and observed pattern for both paralogs and orthologs, the patterns of evolution in these studies are indeed baffling. Further examination of several of these studies has uncovered problems with the analyses such that there is either no longer support for the ortholog conjecture [15] or that there was no statistical support for the ortholog conjecture in the first place, as in the case of human-mouse comparisons in Ref. [2].…”

Evolutionary transition in accessible chromatin landscapes during vertebrate embryogenesis

“…To evaluate transcriptome similarity between samples, a Spearman correlation coefficient was calculated by using the expression values of orthologous genes, as described by Wang et al 8 . As indicated previously 9,79 , phylogenetic relationships between the species being compared were considered in the transcriptome-based identification of vertebrate-conserved stages. For stage combinations among the six different vertebrate embryos (mouse, chicken, softshell turtle, western clawed flog, zebrafish and medaka), we extracted pairs of species that reflected the phylogenetic scale of interest (i.e., vertebrates) and took their average value as expDist.…”