Researchers at the Massachusetts Institute of Technology (MIT) have developed a polygenic scoring model that improves the accuracy of genetic predictions for vastly underrepresented populations, as reported by MIT News. Traditional models, majorly based on studies conducted on those of European descent, exhibit less accuracy for individuals of non-European or admixed ancestry, thereby fostering health disparities.
MIT researchers' model to consider the genetic information of individuals hailing from diverse global ancestries contributes to improvements in health outcomes and advocates for health equity. As per reports from MIT News, there has, on average, been a marked increase in predictive accuracy constituting 60 percent for individuals of African descent and about 18 percent for those bearing admixed genetic backgrounds.
MIT postdoc Yosuke Tanigawa is working on generating a method that enhances accuracy for individuals of admixed and diverse ancestries. This method would ensure that advantages from human genetic research are shared by every individual. Polygenic risk models often demonstrate poor applicability to non-European populations due to varying genetic variation patterns influenced by factors like random drift, demographic history, and environmental elements.
In the course of continuing their work, the researchers strive to include additional data in their model and extend its application to more traits. "This is merely the beginning," expressed Manolis Kellis, a professor of computer science at MIT and a member of the research team.
