Mutation Rate Calculator
Estimate the mutation rate μ from observed de novo mutations, convert per-base-pair rates to mutations per genome and per-year rates, and compare verified rates across organisms.
M = observed de novo mutations · N = callable base pairs · G = generations
Poisson model: P(0) = e-λ, P(≥1) = 1 − e-λ
For RNA viruses, one replication round is the "generation".
Spot-checked against primary literature. "Direct" = mutations counted in pedigrees or mutation-accumulation lines; "phylogenetic" = inferred from species divergence (higher and more uncertain for humans, ~2×10-8).
| Organism | μ per bp per gen | Per genome per gen | Per bp per year | Method |
|---|
What is the mutation rate?
The mutation rate μ is the probability that a single DNA base changes in one generation. It is the molecular clock speed of evolution: small in bacteria (~10-10 per bp), larger in mammals (~10-8), and enormous in RNA viruses (~10-4 per replication round — about one mutation per genome copied, because viral polymerases lack proofreading).
Two honest ways to measure it disagree for humans: counting de novo mutations in parent–child trios gives ~1.2×10-8 (Kong et al. 2012; Michaelson et al. 2012), while dividing human–chimpanzee divergence by an assumed split time gives ~2×10-8. The direct trio number is the default here; the phylogenetic one is the alternative.
Frequently asked questions
What is the mutation rate formula?
The per-base mutation rate is mu = M / (N x G), where M is the number of de novo mutations observed, N is the number of callable base pairs surveyed, and G is the number of generations. For a parent-offspring trio, use the diploid form mu = M / (2 x N_haploid x G). Example: 120 new mutations found across 10 billion callable sites in one generation gives 120 / 10^10 = 1.2 x 10^-8 mutations per base pair per generation.
What does "per base pair per generation" mean?
It is the probability that any single DNA base changes in one generation: roughly 1.2 in 100 million per base in humans. To get mutations per genome, multiply by genome size and ploidy: 1.2 x 10^-8 x 3.2 billion bp x 2 (diploid) = about 77 new mutations per human genome per generation.
What is the human mutation rate?
Direct trio sequencing (comparing parents' and child's whole genomes) gives about 1.0-1.25 x 10^-8 per base pair per generation: Kong et al. 2012 and Michaelson et al. 2012 both land near 1.2 x 10^-8. Phylogenetic estimates from human-chimpanzee divergence give a higher value near 2 x 10^-8, because they depend on uncertain split times and ancestral population sizes.
Why do trio-sequencing and phylogenetic mutation-rate estimates differ?
Trio sequencing measures de novo mutations directly in one generation (about 1-1.2 x 10^-8), but is sensitive to false positives/negatives, callable genome fraction, and paternal age. Phylogenetic estimates (about 2-2.5 x 10^-8) divide sequence divergence by an assumed species split time, so errors in the split date or ancestral polymorphism inflate them. Both methods are honest about different things: this calculator reports trio-based direct rates and notes the phylogenetic value separately.
How many new mutations does a human baby inherit?
About 70 per diploid genome per generation: 1.2 x 10^-8 per bp x 3.2 billion bp x 2 copies = 76.8 expected. The widely cited round number is 60-70, because studies vary in callable genome size and in how they correct for missed or false mutations. Roughly 80% come from the father's germline, and the count rises by about 1-2 mutations per year of paternal age.
What is the difference between mutation rate and mutation frequency?
Mutation rate (this tool) is the tempo: new mutations per base pair per generation. Mutation frequency is a population snapshot: the fraction of individuals (or cells) carrying a mutant allele, e.g. 1 in 10^6. A rate describes how fast new mutations appear; a frequency describes how common mutants are. See our companion mutation frequency calculator for the frequency side.
Is mutation rate the same as allele frequency?
No. Allele frequency (p and q in Hardy-Weinberg) is the proportion of gene copies in a population carrying each allele - a description of what is in the gene pool right now. Mutation rate is the speed at which NEW variants enter the gene pool. Our allele frequency calculator covers Hardy-Weinberg p/q and chi-square tests; this tool covers the mutation rate itself.
What is the E. coli mutation rate?
About 2.2 x 10^-10 mutations per nucleotide per generation, from whole-genome sequencing of mutation-accumulation lines (Lee et al. 2012, PNAS). Because the E. coli genome is only about 4.6 million bp, that is roughly 0.001 mutations per genome per generation - i.e. only about 1 in 1000 cell divisions produces any mutation at all. This is about threefold lower than Drake's classic estimate.
Why are RNA virus mutation rates so much higher?
RNA viruses replicate with RNA-dependent RNA polymerases that lack proofreading, giving typical rates of 10^-4 to 10^-6 mutations per nucleotide per replication (Drake & Holland 1999) - roughly one mutation per genome per replication round for lytic viruses. That is about a million times higher per base than DNA-based organisms, which is why RNA viruses evolve so fast and exist as mutant swarms (quasispecies).
How do you convert a mutation rate between per-generation and per-year?
Divide by the generation time in years: mu_per_year = mu_per_generation / generation_time. Example: human 1.2 x 10^-8 per generation with a 25-year generation time = 4.8 x 10^-10 per bp per year, matching trio-based yearly estimates of 0.38-0.48 x 10^-9. To go back, multiply by generation time. For viruses, one replication round is the "generation."
Should I multiply by 2 for diploid organisms?
Yes, when counting expected new mutations per whole genome: each base exists in two copies, so the expected count is mu x genome_bp x 2. The rate mu itself (per base pair) is unchanged. Yeast, E. coli, and RNA viruses in our reference table are treated as haploid (no x2); human, mouse, fly, worm, and Arabidopsis are diploid.
What makes mutation-rate estimates uncertain?
Four things dominate: (1) false negatives/positives in detecting de novo mutations, (2) how much of the genome is "callable" at high confidence, (3) paternal age - human germline mutations rise with father's age, and (4) whether post-zygotic (mosaic) mutations are counted. Reputable studies report confidence intervals; e.g. the Drosophila estimate of 2.8 x 10^-9 carries a 95% CI of 1.0-6.1 x 10^-9 (Keightley et al. 2014).
Free tool by NRTechStudio