Researchers Develop AI Tool to Enhance Genetic Disease Predictions

A team of researchers in New York has created an innovative artificial intelligence (AI) tool designed to predict genetic risks associated with common hereditary diseases. This development aims to assist healthcare providers in interpreting genetic test results more effectively, ultimately guiding patients toward the appropriate level of care.

Genetic testing reveals variations in a person’s DNA, but many of these changes do not significantly affect health. A single genetic variant seldom tells the complete story, as multiple genes and environmental factors contribute to the risk of conditions such as heart disease and cancer. Addressing this complexity, the research team at the Icahn School of Medicine at Mount Sinai has developed a model that leverages AI and electronic medical records to enhance predictions regarding disease development linked to genetic risks.

Ron Do, a professor of personalised medicine and one of the study’s authors, stated, “We wanted to move beyond black-and-white answers that often leave patients and providers uncertain about what a genetic test result actually means.” He emphasized that by utilizing artificial intelligence alongside real-world lab data—such as cholesterol levels and blood counts—the team can provide better estimates of disease likelihood for individuals with specific genetic variants.

The researchers utilized over one million electronic health records to construct AI models for ten inherited conditions, including breast cancer and polycystic kidney disease (PKD). By applying these models to patients with rare genetic variants, they assigned scores ranging from 0 to 1 to assess the likelihood of disease development. Their findings, published in the journal Science, allowed for the calculation of risk scores for more than 1,600 genetic variants.

The AI model has revealed significant associations between certain genetic mutations and health risks, shedding light on variants previously deemed “uncertain.” Dr Iain Forrest, the study’s lead author, noted, “While our AI model is not meant to replace clinical judgment, it can potentially serve as an important guide, especially when test results are unclear.” He added that doctors can utilize the risk scores to determine whether patients should undergo further screenings or take preventive steps, thus minimizing unnecessary anxiety or interventions for low-risk variants.

Looking ahead, the researchers are expanding their model to encompass additional diseases and genetic variants, as well as a more diverse patient population. “Ultimately, our study points to a potential future where AI and routine clinical data work hand in hand to provide more personalised, actionable insights for patients and families navigating genetic test results,” Do concluded.

This research not only advances the understanding of genetic risks but also offers a path toward more informed healthcare decisions, enhancing the overall management of hereditary diseases.