Fig. 5: Flam-like and unistrand piRNA clusters are somatically expressed.

a Illustration of cell composition in total ovary (germline cells surrounded by somatic follicle cells) or after somatic follicle cell enrichment, respectively. b Soma enrichment library piRNAs mapping to piRNA clusters in sense or antisense orientation across 9 species. The largest cluster representative is shown for species with multiple genome assemblies. For all clusters see Fig. S14a. c Scatterplot of piRNA strand bias against piRNA soma enrichment across the indicated piRNA clusters (9 species). The largest cluster is shown for species with multiple genome assemblies. For all clusters see Fig. S14b. d Genome browser tracks of flamlike1 in D. ficusphila showing transposon annotations (EDTA), soma-enriched piRNAs (cpm), total piRNAs (cpm), soma-enriched RNA expression (ln(cpm+1) scale), and total RNA expression (ln(cpm+1) scale), and mappability. Transposon annotations are shown in red (minus strand) or blue (plus strand). Sequencing data is shown in green or orange for uniquely mapped reads, and grey for multi-mapping reads. e Analysis across de novo identified large piRNA clusters. Clusters were classified as soma, intermediate or germline based on the follicle cell versus total ovary piRNA ratio. Strand bias (log₁₀ scale) is shown across each category. Boxplots show median (central line), interquartile range (IQR, box), and minimum and maximum values (whiskers, at most 1.5*IQR). Data points represent individual clusters. All species shown in Fig. S15. f As in (d) but showing a somatic piRNA cluster in D. virilis. See also Fig. S14e. g Phylogenetic tree summarising all studied somatic piRNA clusters across all analysed species (n = 119). Cluster size represents the mean across all assemblies, except in the “others” category, where a single cluster is shown. Species names in bold have sRNA-seq data to validate their expression. Abbreviations: cpm counts per million, FC follicle cell. Source data are available in the source data file.