A new automated tool called Talos has shown it can lift rare-disease diagnosis rates to 89 percent by systematically reanalysing stored genomic data, according to a Nature Medicine study from June 2026.

Standard clinical interpretation found disease-causing variants in 62 percent of cases. Initial reanalysis raised that to 75 percent. But systematic automated reanalysis with Talos reached 89 percent, identifying 186 of 208 achievable diagnoses.

In a study of 4,735 individuals, the tool found 241 new diagnoses where testing had previously failed. About 32 percent came from new gene-disease links, 22 percent from updated classifications, and 45 percent from variants missed the first time.

The open-source tool was developed by Australia’s Murdoch Children’s Research Institute and Victorian Clinical Genetics Services with the Broad Institute and Microsoft Research. Professor Zornitza Stark said it would transform outcomes for families.

The key insight: many undiagnosed patients already carry answers in their existing data, waiting for better analysis.