Extended Data Fig. 4: Heterochronic shifts in gene regulatory programmes between annelid life cycles.

a,b, Similarity heatmaps showcasing the orthogroup overlap between the clusters of co-regulated genes (see Extended Data Fig. 3c–e), between the three annelids. P-values were derived from upper-tail hypergeometric tests and Benjamini-Hochberg-adjusted. c, Explanation of the orthogroup overlap analysis by quadrants. Clusters were classed as “early” (before dotted lines) or “late” (after dotted lines). Clusters of the female adult of D. gyrociliatus were disregarded. d,e, Heatmaps of relative similarity by quadrants of the orthogroup overlap analyses of the whole genomes (d) and transcription factors only (e). Colour scale in d and e is the same as in c. f, KEGGbrite characterisation of the gene sets under heterochronic shifts (surrounded by dotted black lines in Fig. 2e) between O. fusiformis and C. teleta. g, Bar plots depicting p-values of top biological process GO terms of genes shifted from late expression in O. fusiformis to early expression in C. teleta. P-values were derived from upper-tail Fisher’s exact tests. Full list is available in Supplementary Fig. 13. h, Enrichment analysis of the number of transcription factors per class in clusters of co-transcribed genes of O. fusiformis (left), C. teleta (centre) and D. gyrociliatus (right). For each cluster and class combination, the Bonferroni-adjusted p–value from the two-sided Fisher’s exact test is shown. Cells in red represent overrepresented classes (odds ratio, OR > 1; adjusted p–value < 0.05); cells in blue, underrepresented classes (OR < 1, adjusted p–value < 0.05). Dotted lines highlight clusters of maximal enrichment of the homeodomain class. n.s.: not significant.