Fig. 4

a Schematics of A-to-I RNA editing discovery through sequencing of a parent-offspring trio. The genome sequence of the parents are used as a reference set to distinguish between polymorphisms and editing. b Mismatches between RNA and DNA sequencing data. As RNA libraries were not strand selective, mismatches were read as their complement (i.e., T->C instead of A->G, or C->T as G->A) in roughly half of all cases. c Overlap of editing sites at different time points. The 1.5 and 3.5 hpf samples were more similar to each other than to the 5.3 hpf sample, probably because of replacement of maternal by zygotic transcripts at the MZT. d Association of editing sites with genomic features. A large fraction of RNA editing is classified as “genic_other” due to overlap between introns/exons/UTRs from multiple transcripts. e Number of editing sites in transcripts stemming from different classes of repeat elements. f Number of editing events in individual reads encompassing 10 potential editing sites. The majority of individual reads contained 1–3 RNA editing sites, and never more than five editing sites. Source data: Supplementary Data 1–>3, Source data file.