A report highlighted by the Mohamed bin Zayed University of Artificial Intelligence has renewed attention on the lack of genomic diversity in the datasets used to develop precision-medicine research and artificial intelligence systems. MBZUAI cited an analysis of a major 2023 genome-wide association study database in which 86.5 percent of participants were recorded as having European ancestry, while 0.47 percent were labelled African, excluding African American and Afro-Caribbean categories.
The imbalance matters because genetic associations and biological responses can vary between populations. A risk model trained mostly on one ancestry group may not perform equally well for people whose genetic background is poorly represented. In healthcare, that can affect the usefulness of screening, disease-risk estimation and treatment decisions. It can also make an apparently sophisticated AI system less reliable for the communities that need it to work.
MBZUAI used its Global Genomics Summit to discuss the implications for AI and genomic medicine. The issue is not solved simply by adding more sequences to a dataset. Researchers need representative recruitment, consistent metadata, ethical consent, secure handling of sensitive information and validation in the populations where a model will be used. Clinical professionals also need a clear explanation of uncertainty and limits before a prediction influences care.
The UAE context is relevant because Abu Dhabi is building connections between genomics, healthcare and computational research. MBZUAI’s work includes efforts to combine genomic information with clinical records, biological measurements, lifestyle data and wearable signals. Such multimodal data can help researchers study health in greater detail, but it also increases the responsibility to protect privacy and avoid turning correlation into an unsupported clinical conclusion.
The missing-genomes discussion is therefore not a claim that an autonomous medical system is ready for deployment. It is a warning about the foundations on which AI-enabled medicine is built. Better representation, transparent evaluation and local clinical validation will be essential if precision medicine is to improve outcomes across populations rather than reproduce the limitations of the data from which its models learned. For the Gulf, the opportunity is to contribute regional genomic evidence and governance to a more inclusive global research base.
