Polygenic risk scores (PRS) are promising methods to predict disease risk, but current versions have built-in bias that can affect their accuracy in certain populations and result in health disparities. However, a team of researchers from the Massachusetts General Hospital (MGH), the Broad Institute of Harvard, and Shanghai Jiao Tong University in Shanghai, China, have developed a new approach for producing PRS that better predict disease risk across populations, according to their report inNature Genetics.
Alterations in a genes DNA sequence may lead to a genetic variant that increases the risk of disease. However, most common human diseases, such as type 2 diabetes, high blood pressure, and depression, are influenced not by single genes, but by hundreds or thousands of genetic variations across the genome. Each variant has a significant effect, according to Tian Ge, a PhD in epidemiology, and the co-author of the paper. PRS study the results of genetic variants across the genome and have shown an advantage for clinicians to
Using genomic information from extensive groups of individuals, a PRS must be educated to predict disease risk. However, many disease-causing variations are shared, indicating that the extent to which one gene is linked or is missing in other populations. In some cases, the effect size of a disease, and the extent to which it increases risk, may also vary from one ancestral group to another.
According to Ge, most of the European ancestry genomic research conducted with data was performed, thus creating a Eurocentric bias in existing PRS, resulting in significantly less-accurate predictions and increasing the possibility that they might over- or underestimate disease risk in non-European populations.
Fortunately, investigators have increased their efforts to collect genomic data from underrepresented populations. Using these resources, Ge and his colleagues created a new tool called PRS-CSx that allows them to integrate data from multiple populations and to highlight genetic similarities and differences. Despite the fact that there are still significantly more genomic information on European ancestry, the investigators used computational methods that allowed them to increase the value of non-European data and improve prediction accuracy in ancestrally diverse individuals.
The researchers combined genomic data from individuals from several populations to assess a wide range of physical measures (such as height, body mass index, and blood pressure), blood biomarkers (such as glucose, cholesterol, and the risk of schizophrenia.) The results revealed that PRS-CSx is significantly greater than existing PRS tools in non-European populations.
Ge, who believes that the new approach will continue to be refined in hopes that clinicians may one day use it to inform treatment choices and make recommendations about patient care.
According to the study''s lead author, Yunfeng Ruan, a postdoctoral research fellow at the Harvard University, PRS-CSx might be used, for example, to investigate gene-environment interactions, such as how the effects of genetic risk might depend on the level of environmental risk factors in global populations.
In combination with PRS-CSx, the gap in prediction accuracy between European and non-European populations remains significant. According to Hailiang Huang, the scientist in the Analytic and Translational Genetics Unit at the Broad Institute, and the author of the paper, the expansion of non-European genomic resources will help develop better prediction accuracy in diverse populations.
Ge is an assistant professor of psychiatry at Harvard Medical School (HMS), while Huang is an assistant professor of medicine at HMS.
The National Institute on Aging, the National Human Genome Research Institute, the National Institute of Diabetes and Digestive and Kidney Diseases, the National Institute of Mental Health, the Brain & Behavior Research Foundation, the Zhengxu and Ying He Foundation, and the Stanley Center for Psychiatric Research have all contributed to this effort.