Figure 1.

Transcript switches between cytoplasm and nucleus. (A) The upper panel shows two transcript variants of the gene C16orf91. The lower panel is an example for the identification of a transcript variant switch in the gene C16orf91 using $\Delta TU$. C16orf91-201 and C16orf91-202 were mainly localized in the cytoplasm and the nucleus, respectively, in HeLa-S3 cells (see Supplementary Data 2 for read coverage on C16orf91 gene). P values are calculated with SUPPA2 [32] (see Materials and Methods). (B) Transcriptome-wide identification of transcript variant switches in HeLa-S3 cells. We applied $|\Delta TU|>0.25$ and $p<0.05$ to filter out cytoplasmic ($|\Delta TU|>0$, blue) and nuclear ($|\Delta TU|<0$, red) transcript variants. Unknown (black) are transcripts filtered as no significant. (C) Number of transcript variant switches across the 13 cell lines. The percentage on the histogram indicates the content of protein-coding transcript. (D) Number of genes containing transcript variant switches shared across the 13 cell lines. For example, 4015 genes were identified as being cell specific.