Figure 2.

(A) Evaluation of NMD effect on SOD1 transcripts by comparison of matched tumor exomes and transcriptomes. Visualization with the Integrative Genome Viewer software (version 2.8.4, http://software.broadinstitute.org/software/igv/)³⁵ of WES and RNA-Seq alignments of 3 samples from the Cancer Cell Line Encyclopedia⁵². The missense mutation p.Ile150Val (I150V, sample 3) is used here as a control as it has no effect on NMD activity and is thus found heterozygous in both WES and RNA-Seq alignments. The non-sense mutation p.Glu79* located in exon 3 of SOD1, (E79*, sample 1) was detected on 9% of RNA-Seq reads versus 46% of WES reads, showing the degradation of the SOD1 mutated transcript by NMD. In contrast, heterozygous frameshift mutation p.Lys137Aspfs*26 (K137Dfs*26, sample 2), located in the last exon of SOD1 was detected in 32% of RNA-Seq reads versus 56% of WES reads, highlighting NMD escape. (B) SOD1 mRNA expression correlation with SOD1 mutations in 2029 samples. This plot was generated from the cBio Cancer Genomics Portal (http://cbioportal.org)^17,18. Although not statistically significant, sample 1 with the E79* mutation appears to have a lower SOD1 mRNA expression (z-score = −1.31) compared to sample 2 with the K137Dfs*26 mutation (z-score = −0.21, equivalent to the mean mRNA expression in the wild type group). Deep Deletion indicates a deep loss, possibly a homozygous deletion; Shallow Deletion indicates a shallow loss, possibly a heterozygous deletion; Gain indicates a low-level gain (a few additional copies, often broad); Amplification indicate a high-level amplification (more copies, often focal); Not profiled for CNA indicate the samples for which copy-number analysis was not performed. These levels are derived from copy-number analysis algorithms and indicate the copy-number level per gene.