Harvard-led researchers reported in Nature that ALADYNOULLI can estimate a patient’s lifetime risk of 348 diseases by jointly analyzing electronic health records and genetic data.
683,000-plus participants across UK Biobank, Mass General Brigham and All of Us gave the model up to 52 years of follow-up, letting it track how disease risk changes over time rather than relying on static single-disease scores.
21 disease signatures emerged consistently across the three biobanks, and the model also separated biological subtypes within one diagnosis—such as early- and late-onset heart attacks—pointing to different mechanisms.
23 cardiovascular genetic variants missed by conventional single-disease analyses were identified, while ALADYNOULLI outperformed Pooled Cohort Equation, PREVENT and the Gail model in disease prediction.
The team is now trying to broaden the model, test it in clinical practice and use it to improve patient stratification and clinical-trial design.
How will this AI model ensure fair and accurate health predictions for diverse populations underrepresented in current biobanks?
This AI combines genes and health records. What happens when our daily lifestyle data is added to the prediction?
As AI predicts lifelong disease risk, how can society balance medical progress with an individual’s right to privacy?
ALADYNOULLI Sets New Standard: Predicting 348 Diseases with Interpretable Bayesian Machine Learning and Genetic Integration
Overview
ALADYNOULLI is a new interpretable Bayesian machine learning model introduced by Harvard University researchers. It combines genetic data with longitudinal diagnosis information from electronic health records, allowing it to predict the risk of 348 different diseases. Highlighted in major scientific outlets in 2025 and 2026, ALADYNOULLI represents a breakthrough in healthcare by offering a unified approach to disease prediction. Its ability to integrate diverse data sources provides a more complete and dynamic understanding of how diseases develop, setting a new standard for precision medicine and proactive health management.