Researchers provide a roadmap to get genetic ancestry right for science and society

In a Policy Forum, Anna Lewis and colleagues argue that, for researchers and others who want to invoke genetic ancestry, there is a scientific and ethical imperative to move away from continental ancestry categories and to instead embrace a view of genetic ancestry that reflects continuous variation and historical depth.

Such change is a "prerequisite to any research that looks for connections between genetics and health disparities," the authors say. A continued reliance on continental ancestry categories risks exacerbating medical stereotypes about individuals and groups, contributing to health disparities rather than addressing them, and perpetuating (mis)understandings of race as biological, they add. Many research institutions are reconsidering their use of race as a biological variable and, instead, turning to concepts from genetics to capture differences between groups of humans.

Genetic ancestry – a dominant description of genetic ancestry associated with continents as meaningful groupings – is one of the main proposed alternatives. However, the increasing dominance of the use of continental ancestry categories, such as "African ancestry" or "European ancestry," has led to problematic ethical issues. Here, Lewis et al. highlight the ethical concerns surrounding the continued use of continental ancestry to group individuals. They argue instead for the widespread adoption of a more complex approach to genetic ancestry – a multidimensional, continuous, and category-free concept that reflects our historical depth and the total spectrum of human variation.

Lewis et al. provide a roadmap to achieving this goal, a journey that will require a systems-level change, including new tools, methodologies, and data structures, and a fuller understanding of how and why different fields use and apply the concept of ancestry. "Science is reductive, and a model that uses simple continental categories has been useful in starting the process of understanding human genetic diversity," write Lewis et al. "But all models have their legitimate domains of application and limits, and a much more complex set of models should now be the norm across a wide variety of use cases."