Despite advances in genomic sequencing, up to 80% of patients with rare diseases remain without a diagnosis, often due to undiscovered links between genes and diseases. To address this gap, Queen Mary University of London and University College London researchers developed a novel analytical framework to identify rare variant gene associations in Mendelian diseases.
By analyzing protein-coding variants in whole-genome data from 34,851 cases and their families within the 100,000 Genomes Project, they uncovered 141 new disease-gene associations. Notably, several associations were backed by existing experimental evidence, including links between specific genes and conditions such as monogenic diabetes, schizophrenia, and epilepsy. The research article, published in Nature, demonstrates the transformative potential of large-scale genomic analyses in unlocking new diagnostic possibilities for rare disease patients.
A gene burden analysis framework reveals rare disease genetics
Fewer than half of the 10,000 known Mendelian diseases have an established genetic basis, with many cases likely tied to undiscovered genes. Large-scale initiatives like the 100,000 Genomes Project and other rare disease networks are transforming the field by enabling powerful case-control analyses to uncover these elusive links. These studies provide critical insights into the genetic mechanisms of rare diseases, offering hope for breakthroughs in diagnosis and treatment.
Leveraging data from large-scale sequencing initiatives like the 100,000 Genomes Project, researchers led by first authors Valentina Cipriani, PhD, and Letizia Vestito, PhD, developed an advanced analytical framework to uncover these elusive links. This enhanced tool, called geneBurdenRD, refines statistical modeling and rare variant filtering for Mendelian diseases, enabling the identification of new disease-gene associations.
The framework was applied to 34,851 individuals (72,690 genomes) in the 100KGP rare disease cohort, identifying significant gene-variant associations in 161 analyses across 226 disease categories. Researchers uncovered 165 known and 141 novel disease-gene associations, with novel signals prioritized for further validation based on statistical thresholds and functional evidence. Among these, 69 associations were deemed high-confidence after additional filtering, focusing on genes with supporting functional data or relevance to disease mechanisms.
Cipriani and Vestito reported several key findings on five genes—UNC13A, RBFOX3, ARPC3, GPR17, and POMK—in their Nature article. They identified rare loss-of-function variants in UNC13A that were linked to diabetes with phenotypic overlap. Variants in RBFOX3 were associated with familial generalized epilepsy. Variants in ARPC3 correlated with distinct peripheral neuropathy phenotypes and protein interaction data suggest a connection to known Charcot–Marie–Tooth Disease genes. Rare variants in GPR17 were implicated in schizophrenia with psychiatric comorbidities. Functional studies are needed to clarify the mechanism. Variants in POMK were associated with anterior segment dysgenesis, highlighting potential links to eye development disorders.
Challenges and opportunities in rare disease genetics
The groundbreaking geneBurdenRD framework fills a critical gap in the quest to understand rare Mendelian diseases, offering a tailored approach to gene burden analysis where traditional tools like REGENIE excel in broader, complex genetic studies. By applying this methodology to a large cohort of rare disease cases, researchers have not only reaffirmed known genetic associations but also uncovered numerous previously unreported ones.
However, the study also highlights limitations that point to exciting avenues for future exploration. For instance, while the focus on single-nucleotide variants and small coding region mutations has yielded significant insights, expanding the scope to include non-coding variations and structural variants could reveal even deeper layers of genetic complexity. Additionally, the analysis underscores the challenge of studying ultra-rare conditions—29 of the 226 diseases had fewer than five cases, and 36 fell short of the required sample size for robust statistical testing. This underscores the need for larger-scale sequencing initiatives to capture the genetic underpinnings of rare diseases fully.
The study also advocates for collaborative efforts, such as federated analyses across global rare disease projects. Sharing processed genetic data, ideally through standardized tools like Exomiser, could significantly boost statistical power while adhering to ethical and legal standards. Finally, the researchers emphasize that their work is just the beginning. Future developments must address the intricate genetic landscapes of rare diseases, including factors like incomplete penetrance, variable expressivity, and complex inheritance patterns involving multiple genes.
In sum, geneBurdenRD represents a leap forward, providing a powerful framework to decode the mysteries of rare diseases while inspiring broader collaborations and innovations to tackle the challenges that remain.
