Skip to main content
Fig. 2 | Genome Medicine

Fig. 2

From: Pervasive generation of non-canonical subgenomic RNAs by SARS-CoV-2

Fig. 2

Identification of few prominent dataset-specific non-canonical junctions. a Percentage of junctions that are non-canonical in five independent datasets. Junctions were assigned as canonical if their 5′ location was within 20 bases of the TRS-L and their 3′ location within 15 bases of a TRS-B, and otherwise assigned as non-canonical. Taiaroa, Kim, and Davidson are dRNAseq, while Finkel and Blanco-Melo are Illumina PolyA RNAseq. b Illustration of computational approach to determine the consistency of 5′ and 3′ junction points across independent datasets. The percentage of non-canonical junctions at each genome position in each dataset was determined (X). The mean (μ) and standard deviation (σ) of percentages at each position across the five datasets was calculated. For each position in each dataset, the Z-score was calculated as (X − μ)/σ (i.e., the number of standard deviations away from the mean), and the percentages and Z-scores for each position in each dataset were plotted. c, d Each point represents the percentage and Z-score of one position in one dataset. For each position in the SARS-CoV-2 genome, the percentage and Z-score of non-canonical junctions with a 5′ end (c) or a 3′ end (d) was determined as described above for five independent datasets: Taiaroa, Kim, Davidson, Finkel, and Blanco-Melo. Points with a percentage above 4% of non-canonical junctions and a Z-score above 1 were highlighted

Back to article page