G.AI: an AI-driven platform for phenotype standardization, variant interpretation and structured clinical reporting in rare disease genomic diagnosis
New AI platform cuts rare disease genetic diagnosis time from hours to under an hour while maintaining near-perfect accuracy across thousands of cases.
G.AI is a modular AI platform for rare disease genomic diagnosis that reduces whole-exome analysis from 4-6 hours to ~48 minutes while achieving 95% Top-1 and 99.6% Top-20 variant accuracy in 39,156 multicenter cases. The transparent, traceable workflow addresses scalability and consistency gaps in current manual rare disease variant interpretation pipelines.
What the study was
- Study design
- Multi-center validation; 39,156 WES/WES Trio cases from clinical centers in China; evaluation of Top-1/3/20 variant ranking accuracy, phenotype concordance, and workflow efficiency
- Population
- Patients undergoing whole-exome sequencing for rare disease diagnosis across multiple centers in China
- Sample size
- 39156
- Category
- Diagnostics
- Maturity
- Validated
- Journal
- Journal of Translational Medicine
Why it surfaced
Large multicenter validation (39,156 cases) of AI-driven rare disease genomic diagnosis. Near-perfect Top-3 accuracy (98%) and 5-7x speed improvement directly addresses diagnostic bottleneck for rare diseases. Metabolic disorders achieved 100% Top-3 accuracy. Transparent/traceable design favorable for clinical adoption.
A plain-language summary of published research — not medical advice. Talk to a clinician about your own care.