AI Tool Accelerates Rare Disease Gene Discovery | Hebrew University

AI Tool Speeds Search for Genetic Roots of Rare Diseases

Jerusalem, March 29 – An international research team has unveiled a new artificial intelligence tool designed to dramatically accelerate the identification of genetic causes for rare diseases. The development, announced Sunday by the Hebrew University of Jerusalem, offers a potential lifeline to families facing years of diagnostic uncertainty.

For those affected by rare diseases, obtaining a diagnosis is often a protracted and frustrating journey. Many patients endure years of testing and specialist visits without a definitive answer. This delay not only impacts access to appropriate care but also hinders the development of targeted therapies.

The AI tool, named EvORanker and detailed in the journal Genetics in Medicine, employs a novel approach to gene analysis. Unlike traditional methods that focus on individual genes, EvORanker analyzes genes across a broad evolutionary spectrum. By comparing genetic patterns across more than 1,000 species, the tool can uncover subtle connections between genes and diseases that might otherwise be missed.

In rigorous testing, EvORanker demonstrated a significant improvement over existing diagnostic methods. The AI correctly identified the disease-causing gene as the top candidate in nearly 70 percent of cases and placed it within the top five candidates in 95 percent of cases. This level of accuracy represents a substantial leap forward in the field of rare disease diagnostics.

The impact of this technology extends beyond simply identifying the genetic culprit. Researchers highlighted two specific cases where EvORanker proved instrumental. In one instance, the tool pinpointed a gene responsible for a child’s complex disorder, opening the door to potential treatment options. In another, it elucidated the genetic basis of a multisystem disease, providing crucial guidance for therapeutic interventions.

The potential of EvORanker isn’t limited to diagnosis. Researchers suggest the tool could also facilitate drug repurposing, accelerating the path to treatment by identifying existing medications that might be effective against newly understood disease mechanisms. While the initial focus is on rare diseases, the team is actively exploring applications in cancer research, recognizing the shared challenges of complex genetic landscapes.

The development of EvORanker arrives amidst growing efforts to address the challenges of rare disease diagnosis. According to a recent study published in Nature, a new few-shot learning approach called SHEPHERD is showing promise in diagnosing rare genetic diseases, even with limited patient data. SHEPHERD leverages knowledge graphs and deep learning to perform causal gene discovery and identify patients with similar presentations. This reflects a broader trend toward utilizing artificial intelligence and machine learning to overcome the limitations of traditional diagnostic approaches.

Another AI model, AI-MARRVEL, developed in April 2024, is also accelerating the diagnosis of rare genetic disorders by helping to identify the causative gene of interest. These advancements highlight a surge in AI-driven solutions aimed at tackling the diagnostic odyssey faced by individuals with rare conditions.

For the families grappling with the uncertainty of rare diseases, EvORanker offers a beacon of hope. Researchers emphasize that the tool promises faster, more accurate answers to long-standing medical mysteries, potentially transforming the lives of countless individuals and families worldwide. The next step will involve broader clinical trials and integration into existing diagnostic workflows to maximize its impact.