Fig. 1: Chemosensory tissue transcriptome evolution.

A Overview of the chemosensory tissues and species used in this study. The numbers at the nodes of the species tree are the estimated divergence dates in millions of years. Antennae, proboscis+maxillary palps, forelegs, ovipositors and larvae transcriptomic datasets are labeled in purple, blue, cyan, green and dark yellow, respectively. B PCA of the transcriptomic datasets using 12,096 1:1 orthologs. The percentages on the axes are the amount of variation explained by the PCs. Each dot corresponds to one dataset. C Clustering of the transcriptomic datasets (12,096 1:1 orthologs) according to species and sex. Numbers above branches are bootstrap values for nodes with support <100. Species names are abbreviated to the first three letters. D Relative rate test results arranged by the species' phylogeny. Tissues are arranged to align vertically with C. Colored shapes and lines display the mean and standard deviation of Z-scores from the full set of 1:1 orthologs. D. suzukii, noted with the dashed line and gray font, was used as the outgroup species. Gray data points are Z-scores that resulted from repeating the tests with subsampled datasets (Methods). Asterisks denote the significantly elevated (positive values) or reduced (negative values) rates of gene expression change (Wilcoxon test comparing Z-score distribution to the minimum and maximum values of non-significant Z-scores: dotted lines). Only significant one sample right-tailed Wilcoxon tests are displayed (D. simulans female antennae, n = 4000, p < 2.2⁻¹⁶; D. melanogaster male legs, n = 4000, p = 3.73⁻²⁸⁸; D. simulans larva, n = 4000, p < 2.2⁻¹⁶; D. simulans ovipositor, n = 4000, p = 9.65⁻⁷; D. melanogaster ovipositor, n = 4000, p < 2.2⁻¹⁶) as well as significant one sample left-tailed Wilcoxon tests (D. santomea ovipositor, n = 4000, p = 2.38⁻³³; D. erecta ovipositor, n = 4000, p = 6⁻¹²⁵). Species names are abbreviated to the first three letters. E Fraction of expression tree branches (for all 1:1 orthologs and all tissues) found to best fit one of three evolutionary models: neutrally evolving, constrained, or divergent. Note that the divergent branches were found within otherwise constrained gene trees. Location of source data for this figure can be found in “Source_data.xlsx”.