Automated reanalysis of genomic data for rare disease diagnostics at scale

Scientists developed Talos, a tool that automates the analysis of genetic variants by integrating dynamically updated gene-disease information. The tool was tested on 1,089 individuals with rare diseases and identified 90% of known diagnoses in trio-based analysis. During monthly reanalysis, the number of variants was reduced to one per 200 cases. Application to 4,735 undiagnosed individuals revealed 241 new diagnoses (5.1% yield), with 32% caused by new gene-disease relationships, 22% by new variant-level evidence, and 45% by improved analysis strategies. The automated model enables regular and systematic reanalysis at scale. The study demonstrates that more than 50% of patients remain undiagnosed after initial testing, but reanalysis of genomic data has proven highly effective in finding new diagnoses.