Researchers have identified 34,364 people in Pakistan who carry at least one gene that has been completely knocked out by naturally occurring mutations, across 6,476 different genes. The study, drawn from 173,303 participants in the Pakistan Genome Resource, challenges assumptions about which genes humans truly need to survive. Some of those inactivated genes were previously classified as intolerant to loss of function, raising hard questions about how geneticists define “essential” and what that reclassification means for drug development.
Why Pakistan’s high consanguinity rate rewrites the gene-essentiality map
The core finding is simple but disruptive: thousands of apparently healthy individuals are walking around with both copies of a gene completely disabled. In genetics, a homozygous loss-of-function variant, or homLoF, means neither copy of a gene produces a working protein. The conventional view holds that many such knockouts should be lethal or severely harmful. Yet the Pakistan Genome Resource identified homLoF variants spanning 6,476 genes among its 166,625 exomes and 6,678 genomes.
Pakistan’s relatively high rates of consanguineous marriage, where parents share recent ancestors, make this kind of discovery far more likely than in outbred populations. When parents are related, their children have long stretches of identical DNA inherited from both sides. That raises the odds of receiving two broken copies of the same gene. In populations where such unions are rare, these knockouts stay hidden because most carriers have only one disrupted copy, with the second copy compensating.
This matters because the field’s main yardstick for gene essentiality, the constraint metrics derived from 141,456 individuals in the gnomAD dataset, was built largely from outbred populations of European and East Asian descent. Those metrics flag genes as “loss-of-function intolerant” when they show fewer disabling mutations than expected. But if a gene labeled intolerant turns up knocked out in living, functioning people in Pakistan, the label needs revisiting. The PGR data suggests that some genes scored as highly constrained in gnomAD tolerate inactivation in autozygous backgrounds, where long runs of homozygosity are common. The implication is that constraint scores may overestimate lethality for a subset of genes, and recalculating those scores with PGR allele frequencies could shrink the list of truly essential targets.
The study’s authors, and outside commentators, stress that this is not just an abstract statistical correction. As a news analysis notes, Pakistan’s genetic structure effectively acts as a giant, naturally occurring experiment in gene disruption, offering clues that simply do not appear in more genetically mixed cohorts. The result is a radically redrawn map of which genes humans can live without, at least under the environmental and healthcare conditions represented in the cohort.
From PCSK9 to 6,476 genes: what human knockouts teach drug developers
The practical payoff of finding healthy human knockouts is clearest in drug development. If a person lives a normal life without a working copy of a gene, then a drug that blocks that gene’s protein is less likely to cause dangerous side effects. This logic has already produced real medicines. The most celebrated example is PCSK9: researchers found that people carrying loss-of-function variants in the PCSK9 gene had low LDL cholesterol and protection against coronary heart disease. That discovery directly informed the development of PCSK9 inhibitor drugs now used by millions of patients worldwide.
A similar approach has been applied to LRRK2, a gene targeted by experimental Parkinson’s disease therapies. Large-scale human data showed that people with LRRK2 loss-of-function variants did not exhibit strong adverse phenotypes, helping to evaluate the safety profile of LRRK2 inhibitors under development. The PGR now multiplies this logic across thousands of genes. With approximately 34,000 human knockouts spanning roughly 6,500 genes, the dataset offers a natural experiment on a scale that no clinical trial could ethically replicate.
For pharmaceutical companies, this creates a new shortlist of potential targets. Genes that are fully knocked out in multiple individuals, without obvious early lethality, become attractive candidates for inhibition. Conversely, genes that never appear as homozygous knockouts, even in a highly autozygous population, may move higher on the list of essential functions to avoid. The PGR’s catalogue therefore acts as both a green light and a warning sign, depending on whether a gene is present, absent, or represented only by very rare, possibly deleterious variants.
The study also does not exist in isolation. The Genes and Health cohort, which sequenced 44,028 British South Asians enriched for high autozygosity, has reported similar patterns of biallelic predicted loss-of-function genotypes across thousands of genes. That independent replication in a different South Asian population strengthens confidence that these are real biological findings, not artifacts of sequencing errors or annotation mistakes. Together, these resources suggest that human knockouts are far more common globally than previously appreciated, especially in communities with traditions of cousin marriage.
Gaps in phenotype data and constraint recalculation
The PGR dataset is large, but it has limits that shape how far its conclusions can travel. The most significant gap is the absence of deep, longitudinal phenotype data for the 34,364 individuals carrying homLoF variants. Knowing that someone is alive and enrolled in a biobank is not the same as knowing they are healthy. Some knocked-out genes may cause subtle problems, late-onset diseases, or context-dependent effects that a cross-sectional genetic survey cannot detect. Without linked medical records tracking these individuals over years, the claim that a gene is safely dispensable rests on incomplete evidence.
A second open question involves the mechanics of reclassification. The PGR paper documents homLoF variants in genes previously scored as constrained, but no published tables yet quantify exactly how many of those 6,476 genes were classified as highly intolerant in gnomAD after manual curation. The difference matters. Some apparent knockouts in any large dataset turn out to be false positives caused by sequencing artifacts, misaligned reads, or misannotated splice variants. Others may represent isoform-specific disruptions that spare the most important version of a protein. Until each candidate knockout is carefully reviewed, gene by gene, it is risky to assert that constraint scores are simply wrong.
Recalculating constraint will also depend on how researchers handle population structure. Autozygous segments elevate the chance of homozygous variants, but they also reflect demographic history, such as bottlenecks and founder effects. A gene that appears tolerant in Pakistan might still be essential in other ancestries if compensatory variants, environmental exposures, or pathogen pressures differ. Integrating PGR data into global constraint metrics will therefore require models that separate biological tolerance from demographic quirks.
There is also a social dimension to these limitations. Many of the people represented in PGR live in settings where access to healthcare, diagnostics, and electronic medical records is uneven. Mild or moderate health effects of gene knockouts might go undocumented, especially in rural areas or among poorer households. As a result, the absence of recorded disease cannot be equated with the absence of disease itself. Future work linking genomic data to standardized clinical assessments in Pakistan will be crucial for turning this resource into a reliable guide for precision medicine.
Ethical and global implications
The PGR study underscores the scientific value of including underrepresented populations in genomics, but it also raises ethical questions. Pakistanis have historically been marginalized in global research, appearing mainly as data points in studies designed elsewhere. Now that their genomes are reshaping fundamental concepts like gene essentiality, there is a responsibility to ensure that benefits flow back to local communities in the form of diagnostics, therapies, and capacity-building.
At the same time, the work highlights how cultural practices, such as consanguineous marriage, intersect with biomedical research. While high autozygosity facilitates the discovery of human knockouts, it also increases the burden of recessive disease. Communicating the scientific insights from PGR without stigmatizing communities will require careful engagement with local clinicians, religious leaders, and families. The goal should be to use genetic knowledge to expand reproductive options and improve care, not to pathologize tradition.
For global genomics, the message is clear: relying on outbred, largely European datasets to define what is “essential” in human biology is no longer defensible. As more cohorts like PGR and Genes and Health come online, the catalogue of viable human knockouts will grow, and with it a more nuanced picture of redundancy, resilience, and vulnerability in the human genome. Drug developers, clinical geneticists, and policymakers will all need to adapt their assumptions to this more diverse and empirically grounded view of what humans can live without.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
