Fig 2: Assessment of effects of mutation correction on the transcriptome and genomic landscape. (a) NGS analysis of the on-target site found ABE7.10 corrected 12.4% of mutated alleles whereas ABE8e treatment led to 94.6% correction. Statistical difference and error bars were calculated using the Wald method. (b) 9 of the 10 predicted most likely sites for off-target editing across the genome were interrogated using NGS, illustrating negligible off-target editing activity following ABE7.10 and ABE8e treatment at all 9 sites. Error bars were calculated using the Wald method. (c) 0.5% of EB COL7A1 RNA transcripts contained the nonsense mutation at c.5047 after treatment with ABE8e, compared to 36% without treatment. 3 technical replicates performed. Significance calculated using paired t-test (p = 0.0013) (d) ABE8e treatment led to no increase in A-to-I changes across the transcriptome (e) A heatmap generated from the log2 FPKM of the top 40 differentially expressed genes using the clustermap function of Python’s seaborn library, (version 0.11 https://seaborn.pydata.org/generated/seaborn.clustermap.html) is shown. Untreated EB fibroblasts were compared with fibroblasts treated with 5 µg sgRNA + 2 µg ABE8e (EB-ABE8e), as before, and to fibroblasts treated with 1 µg sgRNA + 2 µg ABE8e. The condition with 1 µg sgRNA + 2 µg ABE8e was termed EB-ABE8e Medium dose and the condition with 5 µg sgRNA + 2 µg ABE8e was termed EB-ABE8e High dose. COL7A1 expression, highlighted in red, was increased after both treatments, with a greater increase following administration of EB-ABE8e High dose.