Table 2 High-throughput sequencing demonstrates specificity of CRISPR mutagenesis
GATK analysisSplit-read analysis
Manually curatedManually curated
MutantReadsPredicted mutationsBoth strainsOne strainNeither strainPredicted mutationsBoth strainsOne strainNeither strain
sy74067,212,7701,419 (151)1465041 (8)22 (2)019 (6)
sy74574,587,8411,441 (173)1673340 (7)22 (2)018 (5)
  • Genomic DNA from two closely related independent dpy-11 mutants was sequenced, generating the indicated numbers of unpaired 50-bp reads, greater than 30× coverage for each strain. Newly arising mutations specific to one strain could be off-target effects of CRISPR-Cas-mediated mutagenesis. The results were compared to the wild-type genome, using the GATK pipeline and split-read analysis. The numbers of mutations predicted by each analysis method are shown; the number in parentheses indicates the number of predicted changes specific to that strain. The predictions were then manually curated by inspection of reads aligned to the reference genome. Of the 1592 changes predicted by GATK, the 324 predicted to be strain specific were analyzed to see whether they could be found in both strains, in only one strain, or in neither. Eight were strain specific; none were likely to be the result of CRISPR mutagenesis (Table S1). All of the mutations predicted by split-read analysis were manually curated; examination of aligned reads could confirm their presence in both strains (24 candidates) or could not confirm their presence in either strain (24 candidates; Table S2 and File S1).