Distal nucleotides affect the rate of stop codon read‐through

Figures

• 

  Figure 1. Examples of ribosomal density profiles with and without SCR. The upper panel shows the ribosomal profile of the A isoform of the Snx1 gene (transcript FBtr0077517), which does not present SCR. The green vertical line in the position zero corresponds to the translation start codon, delimiting the CDS of the 5′ UTR end. The red vertical lines indicate the location of the annotated stop codon (UAA) at position 1376 nt, and a posterior stop codon (UAA) located 129 nt later. The 3′ UTR region after the annotated stop codon (inset) shows the absence of ribosomal reads, as canonical translation termination is efficient. The lower panel shows a ribosomal profile with evidence of SCR in the F isoform of the RpS15Aa gene (transcript FBtr0300828). This example presents an extension after the annotated stop codon (UAG), located at position 393 nt respect to the start codon. The ribosomal density is extended beyond the first stop codon, reaching the second stop codon (UAA) located 189 nt after it (inset).

• 

  Figure 2. An example of phylogenetic conservation of a SCR extension. Ribosomal density profile of the C isoform of the CG13887 gene (transcript FBtr0072583), with a SCR event (top panel). This 30 amino acid extension shows a high conservation level regarding the extensions that correspond to 4 other species of the same genus (bottom panel).

• 

  Figure 3. Frequency of adjacent nucleotides. Frequency of ribosomal leak rate for the nucleotides left‐ and right‐adjacent to the stop codon (left panels and right panels, respectively). The upper panels show histograms associated with the UGA stop codon, the middle ones show histograms associated with the UAA stop codon, and the lower panels histograms associated with the UAG stop codon. The adjacent nucleotides A, C, G and U are identified by yellow, blue, green and red colors, respectively.

• 

  Figure 4. K‐L divergence values analysis of the context sequence. D(r) of the nucleotide frequency usage in position r for the context sequence associated to each stop codon: UGA (A), UAA (B) and UAG (C), indicated at positions 1‐3. Blue and yellow dots correspond to the divergence values calculated for the transcript sets associated with moderate (TS1) and high (TS2) ribosomal leakage rates, respectively. The nucleotide positions in the context sequence are represented in the horizontal axis, while the divergence in its frequency usage is reflected in the vertical one.

• 

  Figure 5. GC3 content in SCR. The fraction of G or C nucleotides at the third position in the codons vs codon position for each stop codon: UGA (A), UAA (B) and UAG (C). Yellow lines correspond to the values calculated for the transcript sets associated with high ribosomal leakage rates (TS2), while the black lines correspond to all transcripts with the same stop codon but excluding the transcripts with SCR. The position corresponding to the stop codons was excluded from the analysis.

• 

  Figure 6. Predictive models for different size context sequences. The top panels illustrate the nucleotide positions in the context sequence, used by three different linear models to predict SCR for the stop codon transcripts UGA, UAA, and UAG, respectively. Black braces indicate the 12 positions considered by [33], and the red ones consider an extended pattern including the 29 positions contiguous to the stop codon. Black arrows consider only 18 positions selected due that they have the highest divergence values associated. The bars in the lower panel show the fraction of erroneous predictions: false positives (pink) and false negatives (light blue), for the three models indicated in the upper panel.

• 

  Figure 7. Ribosomal leak rate estimation. Panel A represents a typical ribosomal density profile, distinguishing the 5′ and 3′UTR ends of the coding region delimited by the AUG start and UGA stop codons (vertical green lines). An enlargement of the stop codon region and the contiguous C‐terminal extension is shown in (B). Here, the ribosomal density accumulated during the 30 nt associated with these regions: δCD and δext is shown (light blue). Panel C shows the location of the ribosome A‐site relative to the UGA stop codon.

• 

  Figure 8. Numerical coding of the nucleotide sequence context. This illustrative diagram represents how the context sequences are encoded numerically to feed the linear model. The * indicate the selected nucleotide positions. Each nucleotide is encoded by a pair of 1 or −1. The procedure is performed for all context sequences and the matrix X is constructed.