AI-Powered Search Unlocks Secrets of “Junk DNA“ to Diagnose Rare Brain Disorders
Saturday, November 22, 2025 – 4:15 PM Update: 22-11-2025 16:18
Researchers at Erasmus MC are pioneering a new approach to diagnosing rare genetic disorders by leveraging artificial intelligence to analyze previously overlooked regions of the human genome. This work,recently published in Cell,focuses on the vast stretches of DNA once dismissed as “junk DNA,” now understood to contain crucial regulatory elements.
Over 8,000 rare genetic disorders affect an estimated 1 to 1.5 million peopel in the Netherlands alone – roughly 6 to 8 percent of the population. Advances in whole genome sequencing have enabled scientists to examine a patient’s entire genetic code, leading to the recent revelation of the rare ReNU syndrome. However,a genetic cause remains elusive for over half of individuals presenting with these conditions.
Customary genetic testing largely concentrates on the 2% of DNA that directly codes for proteins.The remaining 98%, long considered non-coding, is now recognized to house genetic “switches” - enhancers – that control gene activity. Identifying disruptions within these enhancers is key to understanding many genetic diseases, but the sheer volume of non-coding DNA presents a meaningful challenge, akin to “finding a needle in a haystack.”
To overcome this hurdle, a team led by clinical geneticist Stefan Barakat at erasmus MC created a extensive “atlas” of genetic switches active in the brain. They achieved this by mapping the complete genetic material of neural stem cells, the precursors to brain cells, identifying over 140,000 functional switches.
The researchers then developed a predictive AI model, named BRAIN-MAGNET, to prioritize the most impactful DNA components within these switches. BRAIN-MAGNET assigns a score to each DNA building block, reflecting its importance and the potential severity of disease caused by its mutation. This allows researchers to focus their efforts on the most critical areas, accelerating the identification of genetic causes for previously undiagnosed rare brain disorders.
