AI that reads a billion years of evolution shortens the search for rare-disease genes

For families living with a rare disease, the hardest part is often simply getting a name for it. The search for the single genetic mutation behind a child’s condition can drag on for years, through test after inconclusive test — a period doctors call the “diagnostic odyssey.” A new artificial-intelligence tool from the Hebrew University of Jerusalem, published in Genetics in Medicine on April 9, 2026, aims to make that journey dramatically shorter.

The tool, called EvORanker, was developed by Dr. Christina Canavati and Prof. Yuval Tabach and builds on more than a decade of work uniting evolutionary biology and computational science. Rather than relying only on what is already recorded in medical databases, EvORanker looks outward across the tree of life. It compares how genes have appeared, disappeared and changed together across more than 1,000 species, using those deep evolutionary patterns to reveal hidden relationships between genes — including genes never before linked to any disease.

“The search for the single genetic mutation behind a child’s condition can drag on for years, through test after inconclusive test — a period doctors call the “diagnostic odyssey.”

In clinical testing, the approach proved remarkably sharp. EvORanker ranked the correct disease-causing gene as its top candidate in nearly 70% of cases, and placed it within the top five in 95% of cases, outperforming existing tools. In one case cited in the study, it identified a previously unrecognized gene behind a child’s neurodevelopmental disorder after extensive testing had failed; in another, it helped uncover the genetic basis of a severe condition affecting multiple organs.

“Our goal was to give patients and clinicians a tool that can find fast and accurate answers where none existed before,” said Prof. Tabach. There is a second benefit beyond the diagnosis itself: by surfacing new disease genes, EvORanker can point toward existing drugs that might be repurposed to treat them — a shortcut that could save years of development time and reach patients sooner.

The researchers are careful to frame EvORanker as a clinical aid, not a replacement for genetic specialists, and note that further studies are underway to validate its impact in everyday practice. The tool is already available to researchers and clinicians. For the millions of people worldwide affected by rare diseases — most of which are genetic and many of which still have no treatment — a method that can turn years of uncertainty into a faster, evidence-based answer is a genuine and humane step forward.