Mutation Rate Calculator

How to use this calculator

1. Choose the method that matches the experiment. A count of mutant cells alone is not enough to identify a mutation rate.
2. Enter whole-number event and exposure counts in the fields shown. For genomic mode, choose bp, kbp, or Mbp beside callable sequence length; changing the unit preserves the length.
3. For fluctuation mode, enter the number of independent cultures, how many had zero detected mutants, and the mean total viable cells in one final culture. The mean may be fractional but must be at least one.
4. Read the method-specific rate and denominator. Use the result-unit selector to display the same estimate per million opportunities; this changes the scale, not the estimated biology.

Formula and method

For the assumptions behind these results, read the explanation below. You can also explore Mutation Frequency and Generation Time.

Per-locus cases: count new events

At an autosomal locus, each diploid offspring receives two parental copies. In the simple rare-event model, a new dominant phenotype from unaffected parents can represent one newly mutated copy. Dividing qualifying cases by twice the assessed offspring count estimates events per parental transmission at that locus.

Use this approximation only when the phenotype identifies the relevant new heterozygous event, inherited cases are excluded, and ascertainment is complete. Incomplete penetrance, missed cases, parental mosaicism, or other causes of the same phenotype need a more detailed model. The form does not correct those effects. “Per transmission” means one parental copy at the tested locus, not an entire genome.

Genomic rates: define callable exposure

Callable sites are positions where the sequencing and analysis could reliably identify the counted changes. The full genome size can be an inappropriate denominator if some regions were inaccessible. Count independently arising mutation events, rather than every descendant carrying the same event.

Total haploid lineage-generations combines the exposure of all included haploid lineages. Ten comparable lineages followed for 100 generations contribute 1,000 lineage-generations. There is no hidden factor of two: any diploid copies must already be accounted for in the supplied haploid exposure. If callable lengths differ, a proper sum of site-generations is needed; this form assumes the entered length applies across the included exposure.

For example, three new events across 20,000 callable sites and seven lineage-generations give 3/140,000 ≈ 0.00002143 per site per generation, or 21.43 per million sites per generation. This arithmetic illustrates normalization; it is not a typical rate for a particular organism.

Zero-mutant cultures and the p₀ method

The fraction of cultures without detected mutants estimates the probability of no mutation events. Under the model, m = −ln(p₀) estimates events per culture. This calculator uses the m/Nfinal division convention: final population size approximates divisions when the initial inoculum is negligible. It does not divide by the number of population doublings.

For 12 zero-mutant cultures out of 40, p₀ = 0.3 and m ≈ 1.20397. With 200 million final viable cells per culture, the estimate is about 6.02 × 10⁻⁹ per division. The supported interval is 0.1–0.7. Outside it, including all-zero or no-zero cultures, the form requests another method instead of reporting an unreliable rate.

Cultures should begin independently with small mutant-free inocula and reach comparable final populations, with negligible cell death and reversion. This implementation assumes whole-culture detection and does not adjust for partial plating, phenotypic delay, selection, or preexisting mutants.

Rate, frequency, and numerical scale

Mutation frequency is the observed mutant fraction. Mutation rate describes new events relative to a defined opportunity. An early mutation can leave many descendants, so those two quantities cannot generally be substituted for one another.

The result labels preserve the denominator: a locus-copy transmission, a site-generation, or a cell division. “gen” means generation and 10⁶ means one million. A small decimal and its per-million display describe the same estimate. None of these point estimates provides a confidence interval or establishes the cause of an observed difference.