Fig. 5
From: Estimating transcriptome complexities across eukaryotes
Effective Exon Number (EEN) vs Exons per Transcript (EpT) (mean EEN ± 2*SE) in 4 species of animals and three species of plants. A/C) Whole-Transcriptome data and B/D) after conditioning on Orthologs being present in at least one species in the phylogeny. Ortholog data excludes lineage specific genes. EEN is expected to follow a Broken Stick model if intron bounds are randomly drawn from a uniform distribution. Humans and D. melanogaster show lower EEN with a bigger effect at transcripts with high EpT values. Whale sharks show EEN fully consistent with the null expectation. Lancelets show elevated EEN, the highest of any Deuterostome, suggesting more evenly distributed exon sizes than random. While orthologs show a nominal shift in the smallest EpT values, requiring orthology does not alter comparisons to a broken stick model. Hence, Broken Stick Model comparisons are likely to be robust to effects of orthologs and exclusion of lineage specific genes in evolutionary analysis