Fig. 2

Strand expression, sequence bias and secondary piRNA biogenesis signatures of piRNA clusters. a piRNA (25–35 nt) read counts mapping to each DNA strand at piRNA clusters from the assembled chromosomes. b Heatmaps show the information content of the first 18 positions of unique piRNA reads mapping to the positive (upper panel) and the negative (lower panel) DNA strand at piRNA clusters as in a. Sequence logos of the positions with high information content (positions 1 and 10) are shown. c Percentages of piRNA reads having antisense partners at different 5′-5′ or 3′-3′ end overlap distances, from the pool of all reads having antisense partners. d Proportions and nucleotide biases of reads participating in 5′-5′ sense–antisense pairs overlapping with a 10-nucleotide distance, and reads without a 10-bp overlapping antisense partner