Fig. 3

Reproducibility of melRNA-seq. a Normalized read counts (RPM, reads per million 5S RNA reads) of total SINE RNAs in the brain sample and two biological replicates (rep1 and rep2) of spermatogonium samples. b Normalized read counts (RPM) of B1, B2, B3, B4, ID, and MIR families in two biological replicates of spermatogonia samples. c An IGV snapshot showing all reads mapped around a B1_Mur4 locus (inserted in the minus-strand orientation). X indicates reads that were not transcribed from the Pol III transcription start site. d An IGV snapshot showing a poorly expressed B1_Mm locus that is inserted in an expressed gene in the same orientation. The reads with introns likely represent short RNAs produced by mRNA degradation. e An IGV snapshot showing a poorly expressed B1_Mus2 locus having no mapped read in its neighborhood (i.e., stand-alone locus). f Fraction of poorly expressed B1 loci (n = 1464) in terms of isolation of external transcription. g An IGV snapshot showing isolated expression of a B1_Mus2 copy inserted in the minus orientation to the genomic sequence. Read counts were calculated using all reads (i.e., regardless of the presence of a SINE sequence) that were uniquely mapped. Most of mapped reads outside SINEs in this region are mapped to tRNA genes. e Comparison of expression levels of individual B1 loci in the two spermatogonia samples. Pearson’s coefficient R (calculated using RPM values) is indicated. d Power-law distribution of the expression level of B1 loci in the spermatogonia. The histogram shows the loci number for the various expression levels indicated on the x axis. Expression levels correspond to the averages in the two replicates. The most left shows the number of loci where no read was mapped. The values of log2[RPM] ranges from 4 to 16 for expressed loci