AI and Lab Experiments Identify Drug Candidates for Rare Brain Disease
Researchers have identified two existing drugs as potential treatments for Leigh Syndrome, a rare genetic disorder that causes progressive brain damage in children. The discovery combined artificial intelligence screening with laboratory validation, demonstrating how computational methods can accelerate drug development for diseases affecting small patient populations.
Leigh Syndrome affects roughly 1 in 36,000 people. The disease causes motor impairment, intellectual disability, and early death, with few treatment options currently available. The small number of patients and lack of reliable laboratory models have made research difficult.
How the research worked
An international team from the University Hospital Düsseldorf, University of Luxembourg, and CIC bioGUNE in Bilbao used three complementary approaches. They screened compounds using artificial intelligence, tested results in cell cultures and brain organoids derived from patient skin cells, and validated findings in animal models.
The first study examined over 5,500 drugs already approved for other conditions. Sildenafil-currently used for erectile dysfunction and pulmonary hypertension-emerged as a candidate. The researchers used multi-omics analysis to map how sildenafil affected cellular metabolism and function in Leigh Syndrome organoids.
Six patients received sildenafil as a compassionate treatment. They showed preliminary improvements in clinical condition and motor function, according to Prof. Alessandro Prigione of University Hospital Düsseldorf. Larger clinical trials are planned to confirm safety and effectiveness.
A second study used deep learning to screen compounds and identified talarozole, originally developed for acne and psoriasis, as another potential treatment. Researchers filed a patent application for its use in mitochondrial diseases.
Why this matters for rare disease research
Drug repurposing-testing existing drugs for new conditions-reduces development time and cost. Both compounds have extensive safety data from previous use, which streamlines clinical testing.
Prof. Antonio Del Sol, head of computational biology at the Luxembourg Centre for Systems Biomedicine, said computational screening reduces the number of compounds researchers must test in the lab. "With our artificial intelligence expertise, we can build pipelines that help identify novel compounds of interest," he said.
The methodology developed by this team could serve as a template for discovering treatments in other rare neurological disorders. The combination of in silico screening, in vitro testing, and in vivo validation addresses a fundamental constraint in rare disease research: the difficulty of finding enough patients to justify expensive drug development programs.
Learn more about AI for Science & Research and AI for Healthcare.
Your membership also unlocks:
