Open Collections

Dimayacyac, Jose Rafael Santos

University of British Columbia

Phylogenetic comparative methods allow biologists to make inferences about the evolutionary history of phenotypes. Recently, these methods have seen an increase in usage for studying the evolution of gene expression. However, it is unknown whether the distributional assumptions of phylogenetic models designed for quantitative phenotypic traits are appropriate for gene expression data. To evaluate this, I first fit several phylogenetic models of trait evolution to 9 previously published comparative expression datasets, comprising a total of 155,679 gene/tissue combinations. I then assessed the absolute model adequacy for the best fit model of the set. I find that Ornstein-Uhlenbeck models are applicable to 59.8% of genes and 53% of genes were fully adequate across gene expression data sets when using the best fit model. I found that employing species phylogenies allows models to more adequately describe gene expression data than using gene level phylogenies due to phylogenetic error and that widely-used PCMs often fail to account for heterogeneity in evolutionary rate. Using models that allow for multiple rates decreased the amount of statistical inadequacies due to rate heterogeneity. Overall, model adequacy should be carefully assessed when employing PCMs in gene expression evolution studies, especially in regard to phylogenetic error and rate heterogeneity.

Text

2022-12-31

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/83528

Bioinformatics

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/