Identifying the one harmful mutation hidden among tens of thousands in a patient’s genome remains one of medicine’s hardest challenges — especially for people waiting years for a diagnosis.
A new model from Harvard Medical School, called popEVE, could change that.
It doesn’t just flag risky variants — it ranks them by predicted disease severity, even identifying previously unknown genetic causes of devastating childhood disorders.
Clinicians say it could help finally answer the question families have waited too long to hear:
| “What’s causing this?”
But can AI truly transform the rare disease diagnostic pipeline?