BioDeviceHub

Genome Editing Efficiency Calculator

Calculate genome editing efficiency from the fraction of edited sequencing reads or clones, with a confidence interval, or estimate indel frequency from T7E1 band intensities.

Formula

efficiency=neditedntotal\text{efficiency} = \dfrac{n_{\text{edited}}}{n_{\text{total}}}
T7E1:indels %=100(1−1−f),f=cleavedcleaved+uncleaved\text{T7E1:}\quad \text{indels}\,\% = 100\left(1 - \sqrt{1 - f}\right),\quad f = \dfrac{\text{cleaved}}{\text{cleaved} + \text{uncleaved}}
Wilson score interval for the proportion\text{Wilson score interval for the proportion}
ff
fraction of DNA in the cleaved bands of a mismatch-cleavage assay
interval\text{interval}
95% range for the true proportion, allowing for sampling error in the count

How it works

The simplest measure of editing is the fraction of sequenced reads, or of clones, that carry the edit. Because it is a proportion from a finite count, the page adds a 95% Wilson confidence interval: 312 edited reads out of 1,000 supports a true rate of about 28 to 34%. The interval covers counting error only, not bias in amplification or sequencing.

Mismatch-cleavage assays such as T7E1 cut heteroduplexes formed between edited and unedited strands. The fraction cleaved is converted to an estimated indel frequency using the relationship described by Guschin and colleagues. The assay is semi-quantitative, underestimates high editing rates and cannot tell indels apart, so important results should be confirmed by sequencing.

Worked example

312 of 1,000 sequencing reads carry an edit; in a separate T7E1 gel the bands are 620 (uncleaved), 190 and 150 (cleaved).

  1. Sequencing: 312 / 1,000 = 31.2%, 95% CI 28.4 to 34.1%.
  2. T7E1: f = (190 + 150) / 960 = 0.354; indels = 100 × (1 − √(1 − 0.354)) = 19.6%.

Sequencing gives 31.2% and the gel 19.6%. The gel reading is lower, as expected for this semi-quantitative assay.

These are the values the calculator opens with, so you can check its output against this example.

Assumptions

  • Reads or clones are an unbiased sample of the cells edited.
  • For T7E1, the bands are in the linear range of the stain and the cleavage is complete.
  • An edit means what your analysis defined as one.

Common mistakes

  • Reporting a single number without an interval or the number of reads.
  • Treating an indel as a loss of function, which also depends on frame and position.
  • Comparing a gel-based and a sequencing-based rate as though they measured the same thing.

Related tools

Frameshift CalculatorWill an insertion or deletion shift the reading frame?Open calculatorGuide RNA (gRNA) AnalyzerLength, GC content and sequence features of a CRISPR guideOpen calculatorCas9 Target Finder and PAM FinderFind PAM sites and protospacers for SpCas9, SpCas9-NG or SaCas9Open calculatorDNA Translation Tool (DNA to Protein)Translate in any of six reading frames with a chosen genetic codeOpen calculator

Related equipment

Service documentation, failure modes and parts for the instruments this calculation is used with.