The 'Average' Human Genome Carries Millions of Variants, Largest Study Finds

A typical adult's genome departs from the standard human reference sequence at somewhere between 4.3 and 5.3 million sites. That is the central finding of a new survey from the University of Exeter Medical School, which has drawn the most comprehensive portrait yet of what an 'average' human genome looks like. The work rests on whole-genome sequences from nearly 900,000 adults held in two large research biobanks: UK Biobank and the US-based All of Us program.

Professor Caroline Wright, the study's corresponding author, framed the motivation around clinical practice. Whole genome sequencing has become an important tool for diagnosing genetic disease, she said, and medicine is moving toward genetic screening to predict disease. Yet most people, in her view, do not realize how much variation sits in every genome. 'Variation is the norm,' she said, and the study sets out what to expect in the average genome so clinicians and patients can interpret and contextualize results.

Most of those millions of differences are common in the population and fall outside genes, and they are unlikely to have any measurable effect on health. Only about 1.5% of an individual's variants are rare. Ancestry shapes the count, though: people with African ancestry carried about a million more variants than other groups on average. The authors attribute part of that gap to the reference genome itself, which was built mostly from people of European descent.

The picture sharpens when the researchers turn to variants that might matter. The average person carries about 229 variants predicted to disrupt a gene's function, although only about 72% of those were high-confidence calls. Among genes tied to inherited single-gene diseases, the typical person has roughly one to three variants that the ClinVar database classifies as pathogenic or likely pathogenic. Most are single copies of recessive conditions, which means the person is a healthy carrier. Each person also carries about 50 rare 'variants of uncertain significance,' whose effects scientists cannot yet interpret.

The team also counted new mutations, using 2,334 parent-and-child trios. Each person, they estimate, is born with about 61 new mutations in their DNA, plus about two more that arise early in development. Of the new mutations that could be traced to a parent, roughly 75 to 80% came from the father. Paternal age mattered: each added year of age at conception was associated with about 1.2 more new mutations, against about 0.4 for each added year of maternal age.

Polygenic risk scores, which add up the effects of many small variants to estimate disease risk, told a similar story. Across 28 common conditions, the average person ranked in the top 10% for about three of them. About one in five participants of European ancestry scored in the top 1% for at least one condition, as did about one in three of the other participants. The authors think that gap probably reflects the weaker performance of scores developed mainly in Europeans, not real differences in risk.

The takeaway the researchers draw is about context. Everyone carries a small number of variants linked to disease, they found, though many will not actually cause disease in that individual. Genetic test results should therefore be read alongside symptoms, family history, and other clinical and environmental risk factors. The team says the findings should also inform professionals interpreting test results or commissioning genetic screening programs.

The study, by H.I.W. Wright and colleagues, appears in the journal Genome Medicine.