MIPs derive preferentially from abundant transcripts. (A) MIP source transcripts are expressed at a wide range of levels. Expression levels were obtained from Toung et al37 and measured in FPKMs by RNA sequencing. Frequency distribution of expression levels of 15 737 protein-coding transcripts expressed by B-LCLs (black bars) and 1707 (of 1750) transcripts source of MIPs (red line) for which an entry was found in the RNA-seq data. The frequencies of mRNAs and of MIPs source mRNAs are displayed on the left and right y-axes, respectively. Frequencies were calculated for 0.2 FPKM-bin increments. Expression categories (very low, low, medium, and high) were set based on Toung et al37 and confirmed to be unbiased (supplemental Figure 2). (B) The number of transcripts belonging to the expression categories shown in panel A among the 2 sets of transcripts indicate that MIPS derive preferentially from transcripts expressed at moderate to high levels as opposed to transcripts expressed at very low to low levels (*P < 2.2 × 10−6 by Fisher exact test). (C) The 6p21 chromosomal region is a transcriptional hot spot in B-LCLs. Graph shows the number of transcripts belonging to the expression categories shown in panel A among 316 protein-coding transcripts located in the 6p21 chromosomal region and the protein-coding transcriptome of B-LCLs (*P < .008 and **P < 2.05 × 10−7 by Fisher exact test). The average expression level (FPKMs/gene) is shown on the top for both datasets. The 6p21 region analyzed was located between positions 30400001 and 46200000 (UCSC Genome Browser, http://genome.ucsc.edu/; and NCBI Map viewer, http://www.ncbi.nlm.nih.gov/projects/mapview/). (D) The 6p21 transcriptional hot spot is a preferential source of MIPs. Comparison of the number of MIP-coding transcripts located in 6p21 (6p21+) and in the rest of the genome (6p21−; *P = .013 × 10−6 by Fisher exact test).
