Fig. 7: Several Gypsy transposons are exclusively regulated by flamlike1 and flamlike3.

Genomic origin of piRNAs that are antisense to transposons in D. melanogaster (a, b), D. ficusphila (c, d), D. persimilis (e, f), and D. pseudoobscura (g, h). Barplots (right) display the number of uniquely mappable piRNAs against each TE in soma-enriched (a, c, e, g) or whole ovary (b, d, f, h) samples. The TEs are arranged in decreasing order following their somatic-to-germline enrichment. The number of piRNAs that map to the indicated clusters are coloured according to cluster strand (sense, blue; antisense, red) and piRNAs mapping elsewhere in the genome are shown in grey. Labels are shown for subfamilies that are exclusively controlled (>90% of piRNAs) by a cluster and best hit to known TEs are indicated if available (80/80/80 rule). Boxplots (left) summarise the fraction of piRNAs antisense to individual transposons derived from each cluster. Total cluster-derived piRNA abundance (white) are further subdivided into the sense (blue) and antisense (red) cluster strands. The number of transposon subfamilies covered by each cluster (>10 reads) are indicated under each boxplot. Boxplots show median (central line), interquartile range (IQR, box), and minimum and maximum values (whiskers, at most 1.5*IQR). Pooled counts from 2–4 biological replicates. Abbreviations: TE transposon, cpm counts per million. Source data are available in the source data file.