Most people carry two copies of each gene, but across a whole population some versions of a gene are common and others are rare. Allele frequency is a measure of how common one gene version is in a group, not in a single person. It stays steady unless something pushes it: chance, migration, natural selection or new mutations. Those frequencies are a record of a population’s history, shaped by luck, movement and what helped ancestors survive. So the same variant can be ordinary in one part of the world and almost absent in another, which affects how doctors read DNA test results.
How Allele Frequency Is Counted
A gene version is called an allele, and the share of copies in a group that are a particular allele is its allele frequency. The counting is by copies, not by people. Since each person has two copies of most genes, 100 people carry 200 copies. If 50 of those 200 copies are one version, its frequency is 25%, or 0.25.
Take a classroom of 10 people and a gene with two versions, B and b. That’s 20 copies. Say three people carry two copies of b, two carry one of each, and five carry none. The b copies number 3×2 + 2 = 8, so b has a frequency of 8 out of 20, or 0.4. Genotype frequency counts people by the pair they carry instead: 30% have two b, 20% have one of each, 50% have none.
Every copy of the gene in the group is one version or another, so the shares always total 100%, or 1.0. If you know all but one, you find the last by subtraction. That’s why a gene with two common versions can be summarized with a single number: a version at 0.25 means the other sits at 0.75.
The share of people who carry at least one copy is usually higher than the version’s frequency, because carriers include people with one copy and people with two. A version at 10% of copies can be carried by nearly 19% of people when partners pair up at random, because most copies sit one per person. News stories sometimes say “X% of people have this variant” when they mean carriers, and sometimes when they mean copies, so check which one is being described before comparing numbers.
Frequencies also describe a group only. A 25% frequency doesn’t mean you have a 25% chance of anything personally. A variant can be rare overall but common in your family, and what you inherited depends on your parents, not on the population average.
The Baseline When Nothing Changes
In a large population where people pair up at random, gene versions keep the same frequencies from one generation to the next.1Nature Education. Hardy-Weinberg equilibrium Inheritance shuffles copies without creating or destroying them. This baseline is called Hardy-Weinberg equilibrium, and it holds only if there is no mutation, no migration, no natural selection and no chance swings in small groups.2National Institute of Justice. Population Genetics and Statistics for Forensic Analysts Random pairing means people of all genetic types have children together equally, and no selection means every type survives and reproduces equally well.3LibreTexts Biology. Population Genetics
Real populations always break at least one condition, but the baseline is still useful for spotting a tilt. Forensic labs check their reference data against it before using it.
Real people don’t pair randomly for genes. They often choose partners from nearby, from similar backgrounds or with shared community ties, which shifts how gene versions are combined in children. For most gene versions, such as those with no effect on looks or personality, nobody is choosing partners based on them, so the assumption is close enough to work well. Traits people do choose on, such as height, can depart further from the baseline.
Inbreeding does not change allele frequencies by itself, but it can change genotype frequencies a lot.4University of Washington. Understanding Allele Frequency and Distribution in Genetics As relatives have children together, more people end up with two identical copies while the overall share of each version stays put.5Plant Breeding and Genetics. Introduction Selection then acts differently, since it works on the pairs people actually carry.6University of Washington. Understanding Allele Frequency and Distribution in Genetics
The Forces That Make Gene Versions Rise and Fall
Mutation supplies new versions. Each child is born with roughly 150 brand-new DNA changes not present in either parent, within a range of about 100 to 200.7Nature. Human de novo mutation rates from a four-generation pedigree reference About 70 to 80% of those new changes come from the father, and the number rises with the father’s age.8Nature. Human de novo mutation rates from a four-generation pedigree reference Every new allele therefore starts out extremely rare, as a single copy in one person.
Most new changes fall in parts of DNA that do nothing noticeable, so they are neither helpful nor harmful. A few change a gene’s working and can be harmful, and very rarely one proves useful. Because each new version begins as a single copy, most are lost within a few generations purely by luck, even when harmless.
Genetic drift is change by luck: which copies happen to get passed on. Ten coin flips can easily give seven heads, but a thousand flips rarely give 700. A small village is like ten flips, and a version can vanish by luck, which is why drift reduces genetic variety most strongly in small populations.9PubMed Central. Effects of Genetic Drift and Gene Flow on the Selective Maintenance of Genetic Variation In large populations luck evens out.
If a version is carried by only a few people in a small group, it can disappear when those people happen to have no children or don’t pass that copy on. Once lost, it can’t come back through drift. Only a new mutation or the arrival of migrants can restore it. That’s why small isolated populations tend to become more alike inside, with fewer versions of many genes.
Gene flow, when people move and have children with another group, mixes gene versions between populations. High enough gene flow makes two groups so similar they behave as one population.10Wikipedia. Gene flow Over the last 5,000 years in Europe, migration and mixing changed allele frequencies more than natural selection did.11PubMed Central. The contribution of gene flow, selection, and genetic drift to five thousand years of human allele frequency change
Natural selection raises versions that help survival or reproduction in a given environment. Even so, selection-driven change is uncommon across most of the human genome.12PubMed Central. The contribution of gene flow, selection, and genetic drift to five thousand years of human allele frequency change
Why Some Populations Carry Variants Others Lack
In Europe, the ability to digest milk as an adult comes mainly from a single gene version.13PubMed Central. Evolution of lactase persistence: an example of human niche construction That version spread under strong selection over the last 5,000 years,14PubMed Central. Diversity of lactase persistence alleles in Ethiopia: signature of a soft selective sweep and its age matches the start of herding animals and dairying.15PubMed Central. Evolution of lactase persistence: an example of human niche construction In Africa and the Middle East, several different versions produce the same trait.16PubMed Central. Evolution of lactase persistence: an example of human niche construction Different populations reached the same advantage by different genetic routes.
Most adults worldwide lose much of their ability to digest milk sugar after childhood, which is the original, ordinary state of the gene. People without the persistence version may get bloating, gas or cramps after milk, and many handle small amounts or fermented foods like yoghurt and hard cheese better. Being intolerant is not a disease but one common version of a trait.
When a population starts from a few people, the versions those founders carried can become unusually common. A single disease-causing variant can then dominate a whole group, a pattern called a founder effect.17PubMed Central. The prevalence, genetic complexity and population-specific founder effects of human autosomal recessive disorders Strong founder effects show up in 29 recessive conditions, including cystic fibrosis.18PubMed Central. The prevalence, genetic complexity and population-specific founder effects of human autosomal recessive disorders
Carriers with one copy of a recessive variant are healthy and unaware. In a small founding group, carriers are more likely to have children together, so the condition becomes more common. This is why some communities are offered specific carrier screening.
Geography creates gradual gradients too. Species often spread across changing temperature, rainfall or day length, and different settings favour different versions.19Wikipedia. Cline (biology) A trait can change gradually across a region.20Population Ecology. Mechanisms and tests for geographic clines in genetic polymorphisms That’s one reason skin colour or milk tolerance doesn’t map neatly onto race: most variation lies within groups, not between them, gradients cross borders gradually, and the same trait can have different genetic causes.
What Allele Frequency Tells Us About Disease
Large studies have found hundreds of common variants linked to diseases like diabetes and heart disease, but each adds only a little to risk.21PubMed. How important are rare variants in common disease? Over 90% of the inherited part of most common diseases remains unexplained by these common variants.22PubMed Central. Common vs. Rare Allele Hypotheses for Complex Diseases Researchers think rare variants, often in many different genes, may matter more than once assumed.23PubMed. How important are rare variants in common disease?
For someone reading a risk result, a common risk variant usually shifts odds only slightly. A person with many of them might have somewhat higher or lower odds than average, not a certain outcome. Diet, activity, smoking, age and other factors also affect diseases like diabetes and heart disease, and often matter as much as or more than these variants. Family history may capture risks no test found. A rare high-impact variant is a different matter and usually warrants a genetic counsellor.
Rarity itself can be a clue. Variants that cause serious disease are kept rare by selection, so a very common variant is unlikely to cause a severe childhood disease. Labs lean on exactly that logic when they use frequency databases.
Allele Frequency on a Genetic Test Report
Your report may cite gnomAD, a large public database of how common gene versions are, used widely in clinical labs.24PubMed Central. Improved allele frequencies in gnomAD through local ancestry inference It lists frequencies separately for African/African American, East Asian, European, Latino, South Asian, Ashkenazi Jewish, Finnish and Middle Eastern groups.25PubMed Central. Variant interpretation using population databases: Lessons from gnomAD A variant rare overall can be common in one group, so the group-specific number matters. A lab using only the overall number might wrongly label such a variant suspicious. The reverse happens too: a variant common in a well-studied group may be reassuring, while the same variant in an underrepresented group is judged against thin data.
The lab is checking whether the variant is too common among healthy people to plausibly cause a rare disease. If a variant that supposedly causes a severe rare condition appears in many healthy people, that is strong evidence it is harmless. If it is absent or extremely rare, that doesn’t prove it is harmful. It only fails to rule it out.
A “variant of uncertain significance” means the lab couldn’t decide whether the variant causes disease. Frequency data is one piece of evidence used, and results can be reclassified as databases grow.
It’s reasonable to ask which population frequency was used. If your ancestry is underrepresented, mention it. You can also ask whether relatives should be tested.
How DNA Identification Uses Allele Frequencies
Forensic DNA profiles look at around 20 to 24 marker spots, each with many common versions.26Pomeranian Journal of Life Sciences. Allele frequency distribution and forensic parameters of 23 autosomal STR loci in the Polish population using the PowerPlex® Fusion 6C System Each spot is inherited largely independently of the others, so the frequencies at each spot are multiplied, and combined across 23 spots the chance two unrelated people match is vanishingly small.27Pomeranian Journal of Life Sciences. Allele frequency distribution and forensic parameters of 23 autosomal STR loci in the Polish population using the PowerPlex® Fusion 6C System A pair of versions that is common at one spot makes only a small cut in the odds, but many spots multiplied together shrink the number rapidly.
Those odds depend on using frequencies from the right population.28PubMed Central. Autosomal str allele frequencies, forensic parameters and population structure in four underrepresented indigenous groups from Paraguay Most variation lies within groups, yet using a blended national dataset for a distinct community can mislead.29PubMed Central. Autosomal str allele frequencies, forensic parameters and population structure in four underrepresented indigenous groups from Paraguay A version that is common in one community and rare in another gives a very different match probability depending on which table is used, and broad figures can make a match look rarer than it really is. So labs build reference data for specific communities, such as Indigenous groups in Paraguay or a tribal community in India.30PubMed Central. Autosomal str allele frequencies, forensic parameters and population structure in four underrepresented indigenous groups from Paraguay
A “one in a billion” figure in court is the chance that a random unrelated person matches. Relatives share more versions, so the odds are weaker for them. The figure says nothing about how the DNA got there.
Identical twins are an exception. They share essentially the same DNA, so a profile can’t normally tell them apart.
Why Genetic Variety Matters for Wildlife and Crops
As an endangered species shrinks, it loses gene versions, and they don’t come back when numbers recover.31UC Museum of Paleontology. Relevance of Evolution New variety only accumulates through mutations over many generations.32UC Museum of Paleontology. Relevance of Evolution Small, low-variety populations are more vulnerable to extinction and inbreeding.33NOAA. Preserving Genetic Diversity Gives Wild Populations Their Best Chance at Long-Term Survival Crops also carry less variety than their wild ancestors, especially in genes chosen during domestication.34ACS Publications (Journal of Agricultural and Food Chemistry). Genetics and consequences of crop domestication
Headcount alone can mislead. A population that has recovered in numbers may still carry little variety if it passed through a very small stage, so it can look healthy while remaining vulnerable. Such a group may cope poorly with a new disease or a changing climate because few versions give it options.
Conservationists respond in several ways. Moving animals between isolated groups restores gene flow, and breeding programs track genetic variety, not just headcount. Seed banks keep wild relatives of crops, which carry versions, such as resistance to pests or drought, that modern varieties lost. Breeders can cross these back in when a crop faces a new disease.