Some people can live apparently ordinary lives even when both copies of a gene (one inherited from the mother and the other from the father) have stopped working. Until their DNA is sequenced, they may never know that this has happened. Each such individual provides researchers with something close to an experiment that nature has already performed.
A large study from Pakistan, published in Nature (June 2026), has now identified such naturally occurring ‘human knockouts’ on an unprecedented scale. Among 1,73,303 participants, about one in five carried at least one gene in which loss-of-function variants had disrupted both copies. Altogether, the study identified such knockouts across 6,476 genes, which is close to one-third of the roughly 20,000 protein-coding genes in humans.
The study also demonstrated the value of ‘recall by genotype’. Instead of merely recording a genetic variant, researchers could identify a participant with an unusual genotype, contact that person again, study relatives and perform specialised tests designed around the affected gene. One important example was APOC3. The APOC3 gene produces a protein involved in the metabolism of triglyceride-rich particles. Participants lacking functional copies of APOC3 had substantially lower fasting triglyceride concentrations and a smaller rise in triglycerides after consuming a fat-rich meal. This provided human evidence that lowering APOC3 activity could be therapeutically useful. Human genetics can also warn researchers away from ineffective targets. For instance, it was found that the loss of the gene PLA2G7 substantially reduced Lp-PLA2 levels (high levels of which are associated with an increased risk of coronary artery disease), but did not reduce coronary disease. A drug called darapladib, designed to inhibit the same enzyme, had already failed to improve outcomes in large cardiovascular trials. Such genetic evidence helps distinguish promising drug targets from attractive biological theories that may not translate into clinical benefit. They linked the genetic information to medical histories and laboratory biomarkers. This is an important reminder that global genomic knowledge remains incomplete when large populations are poorly represented. Genes important for embryonic development, DNA repair and energy production were conspicuously resistant to knockout. The clinical implications go beyond rare genetic diseases. Loss of the gene CIDEB, for example, was associated with lower liver enzyme concentrations and a lower risk of fatty liver disease, making the pathway potentially interesting for drug development. Other findings offer warnings. The gene LRRK2 is an important therapeutic target in Parkinson’s disease. The Pakistani data suggested that loss of the gene may be associated with impaired kidney function, raising questions about whether prolonged pharmacological inhibition could have renal effects. The resource can also challenge assumptions derived from animal experiments. Mice lacking PRDM9 are infertile, yet Pakistani men and women with loss of function in the same gene had children. Biology observed in laboratory animals does not always predict what happens in humans. Nor should the scientific value of these studies be interpreted as an argument in favour of consanguinity. Marriage between relatives increases the risk of recessive genetic disorders. Research involving such populations requires informed consent, privacy protection, genetic counselling and fair sharing of benefits. The next step should go beyond cataloguing DNA. A reference map records which variants exist; it cannot say what they do. For decades, scientists have learned what genes do by deliberately switching them off in flies, mice and laboratory cells. Human knockout studies reverse that approach. Researchers first find people in whom nature has already switched off a gene and then ask what happened. Studied carefully, ethically and inclusively, it can help explain disease, identify safer drug targets, prevent expensive therapeutic failures and bring medicine closer to understanding what individual genes actually do in the human body.
That requires linking genomes to medical records and laboratory measurements, and consent to return to participants, so that someone found to be missing a gene can actually be examined. Human genetic diversity is therefore more than a record of our ancestry. The value of this approach became clear in a 2017 Nature study led by Danish Saleheen. Researchers analysed the protein-coding regions, or exomes, of 10,503 adults enrolled in the Pakistan Risk of Myocardial Infarction Study. They identified 1,843 participants carrying variants affecting both copies of at least one gene, involving 1,317 genes. These genetic findings were compared with more than 200 biochemical and disease-related traits. Several years later, olezarsen, a medicine designed to reduce APOC3 production, was approved in the United States in December 2024 for familial chylomicronaemia syndrome, a severe inherited disorder of triglyceride metabolism. The Pakistan Genome Resource, reported in June 2026, expanded this approach enormously. Researchers analysed data from 1,73,303 participants across 23 cities, including 1,66,625 exomes and 6,678 whole genomes. About 6.6 million coding variants were identified. Nearly 47% had not been observed among non-South-Asian individuals in gnomAD, a major international genetic variation database. Around 30% were absent even when South Asian data were included. Across the Pakistani cohort, 34,364 people carried a predicted loss-of-function variant affecting both copies of at least one gene. In total, 6,476 genes were represented by at least one such individual.
