Figure 1

Identification of genes associated with long upstream antisense transcripts in DN thymocytes. A) Heatmap showing the Total RNA-seq signal from ΔRag (DN) thymocytes (SoliD platform) found in the 5 kb region surrounding the TSS of all non-overlapping Refseq genes. Signal was computed based on the number of reads per 100 bp binned regions originated from either the antisense or sense strand with respect to gene orientation (left and right panels, respectively). The heatmap is ordered according to the antisense signal for the [-5 kb; 0] region. The threshold for significantly expressed antisense transcripts is shown by a dotted line (see Methods section). B) Examples of genes associated with LUAT in ΔRag thymocytes. The Total and PolyA RNA-seq signals for the plus and minus strands are shown. Arrows indicate transcript orientation. The scales and genomic coordinates are shown on the left and top of each panel, respectively. Note that the scales were independently fixed for the plus and minus strands in order to properly visualize sense and antisense transcripts. C) Average profiles of PolyA- RNA-seq signal in ΔRag thymocytes for LUAT-associated genes (red line) and a control set of similarly expressed genes (black line). Signals corresponding to the orientation of the coding genes are represented as positive values while antisense signals as negative values. D) Histogram of the positions of 5′ end of LUATs relative to the TSS of their associated coding-genes. E) Distribution of expression of all coding genes (red), LUAT-associated genes (green) and LUATs (blue) in ΔRag thymocytes. F) Number of LUAT-associated genes in each expression quartile of all coding genes (Q1 = 3.05e-6 FPKM; Q2 = 0.013 FPKM; Q3 = 1.99 FPKM). The red line indicates the expected (random) distribution.