New Long-Read DNA Test and AI System Advance Rare Disease Diagnosis
Researchers recommend a new long-read DNA test as the first choice for rare genetic disorders, showing a 3% higher diagnosis rate. Separately, a Chinese AI diagnostic system called DeepRare achieved 57.18% accuracy using only patient symptoms, with accuracy exceeding 70% when genetic data is included.
A new long-read genome sequencing test provides a more complete picture of DNA and leads to a diagnosis more often for patients with rare genetic disorders, while a separate Chinese research team has developed an AI-powered diagnostic system that sets a new record for accuracy. These represent two significant recent advances in the challenging field of rare disease diagnosis.
Researchers from Radboudumc and Maastricht UMC+ compared standard diagnostics with the new DNA test in a study of 1,000 patients. They found the long-read sequencing test yields three percent more diagnoses and can replace fifteen other tests. "We recommend using this test worldwide as the first choice," said Professor of Translational Genomics Lisenka Vissers.
The test is based on long-read genome sequencing, which reads DNA segments of up to twenty thousand building blocks, compared to the current standard of fragments of about three hundred. "Like a jigsaw puzzle, assembling the DNA puzzle is much easier with such large pieces, resulting in a more complete picture," explained Professor of Genome Bioinformatics Christian Gilissen. The test also reads modifications on the outside of DNA that can switch genes on or off, which sometimes cause rare disorders. "With current diagnostics, this requires additional specialized tests, but with long reads we capture these modifications as a bonus—two in one."
The research, published in the New England Journal of Medicine, notes that while a condition is considered rare if it affects fewer than one in two thousand people, up to 400 million people worldwide have a rare disease. Eighty percent of these have a genetic cause, and diagnosis often takes years.
Meanwhile, a Chinese research team has developed an AI-powered rare disease diagnostic system called DeepRare. The system was developed by a team from Xinhua Hospital, affiliated with the Shanghai Jiao Tong University School of Medicine and School of Artificial Intelligence. The study was published in the journal Nature.
Test data show that when only patients' clinical phenotypic information was provided without genetic data, DeepRare achieved a first-attempt accuracy of 57.18 percent in phenotypic diagnosis—an improvement of nearly 24 percentage points over the previous global model. When genetic data were incorporated, its diagnostic accuracy exceeded 70 percent.
Since its online diagnostic platform launched last July, DeepRare has registered over 1,000 professional users across more than 600 medical and research institutions worldwide. The system integrates real-time access to a vast repository of medical literature knowledge and real-world clinical case data. In terms of diagnostic reasoning, it employs an iterative cycle of hypothesis, verification and self-reflection. Each diagnostic conclusion comes with a complete chain of evidence, allowing doctors not only to see the diagnosis but also to understand the underlying basis.
The research team is preparing to initiate a global AI alliance for rare disease diagnosis and treatment. They plan to complete real-world validation of 20,000 rare disease cases within the next six months.