Genomic Medicine · global
Carrying a pathogenic variant does not mean having a disease: Study of over 900,000 genomes establishes a reference for interpretation
A study spanning two major biobanks in the UK and US found that a typical individual genome carries about 1 to 3 rare variants classified as pathogenic or likely pathogenic, most consistent with carrier status for recessive diseases. As sequencing increasingly moves into health screening, understanding these findings is more important than identifying the variants themselves.
When the word “pathogenic” appears in a genetic test report, readers can easily interpret it as a prediction of disease. However, a new study analyzing the whole genomes of over 900,000 people shows that carrying disease-associated variants may simply be part of an individual’s genetic makeup. Once a variant is detected, a more important question remains: Will it cause disease in this person?
A team led by the University of Exeter in the UK published the study in *Genome Medicine* on October 22, analyzing 490,541 participants in the UK Biobank and 414,830 in the US All of Us program, and comparing six genetically inferred ancestry groups. The “average genome” described by the study is a reference range for how many variants, and which types, an individual typically carries, rather than a standard sequence that everyone should match.
Across a curated set of genes associated with single-gene diseases, a typical individual genome carries about 1 to 3 rare variants classified by the ClinVar database as “pathogenic” or “likely pathogenic.” Crucially, most are heterozygous variants in recessive disease genes, meaning that only one of the two gene copies carries the variant, consistent with carrier status. A pathogenic classification in a database therefore cannot be directly translated into a disease diagnosis for the person carrying the variant.
In addition, each person has about 50 rare variants of “uncertain clinical significance”: the available evidence is insufficient to determine their clinical impact. These figures provide context for test results and serve as a reminder that sequencing can identify differences in DNA, but this does not mean medicine can already explain the consequences of every difference.
Another layer of risk comes from the combined effects of many variants. The team analyzed polygenic scores for 28 common diseases in the UK Biobank and found that, on average, each person had scores in the top 10% for about three diseases. But score rankings describe relative genetic risk; they cannot be interpreted to mean that an individual will inevitably develop three diseases, nor do they establish that person’s actual probability of developing them.
An important limitation of the study is that the team did not examine participants’ health status. Both the University of Exeter’s news release and the report supplied by the university to Medical Xpress emphasized that these results cannot establish that people carrying the relevant variants already have a disease. Even someone carrying a pathogenic variant may not develop the disease because of incomplete penetrance; interpretation still requires consideration of symptoms, family medical history, and other clinical and environmental factors.
This reference range also has limitations in population coverage. Despite the large sample, the data remain skewed toward European ancestry, with many populations worldwide underrepresented. As whole-genome sequencing moves from disease diagnosis into screening, the study offers a starting point for understanding results: identifying meaningful risks while avoiding prematurely turning genetic differences commonly found in populations into conclusions about an individual’s disease status.