New software tool facilitates diversity in genetic studies


Genome-wide association studies (GWAS) have generally excluded diverse and minority individuals in the search for genetic variants that confer disease risk.

Researchers at Harvard-affiliated Massachusetts General Hospital (MGH), the Broad Institute at MIT and Harvard, and other institutions around the world have now developed an open-access software package called Tractor that increases the discovery power of genomics in poorly studied populations. A study of the performance and accuracy of the tractor has been published in Genetics of nature.

Researchers perform a GWAS to identify where the genetic variants responsible for the disease are found in the genome. Recently, geneticists have started to create models from published GWAS data to predict disease risks in individuals. But the clinical utility of these models is currently limited, as most are based on genomic studies of people of European descent.

“If you build disease risk models on the available data and attempt to extrapolate them to various populations, the accuracy of predicting who will get sick is reduced,” said Elizabeth Atkinson, lead author of the paper and researcher in Analytic and Translational Genetics. Unit (ATGU) at MGH. “These errors exacerbate existing health disparities, in part because we cannot find specific genetic variants that could contribute to a higher risk for a particular disease in various populations.”

Another significant drawback of current GWASs is that they “leave a lot of genetic discovery opportunities on the table for all populations,” Atkinson said. People of African descent, for example, have on average a million more genetic variations than a person who is not of African descent due to human migration patterns through the ages. Conducting a GWAS with diverse populations allows geneticists to identify genetic associations to the disease in many other places in the genome, Atkinson said.

“Within these genomic regions identified in a GWAS, the genetic mutation that actually causes the disease is mostly shared among the ancestors,” she added. By studying mixed populations – people of recent ancestry belonging to two or more previously isolated population groups, such as Africa and Europe – “we can get stronger and more precise genetic association signals and make a better job of locating the causal mutation, which improves our understanding of the disease for everyone.

Until now, there has been no precise way to control the composition of ancestors in the mixed groups studied in a GWAS. “Different ancestry groups have genetic variants that occur at different frequencies due to the demographic history of populations,” Atkinson explained. “Failure to account for ancestry in a GWAS can lead to false positive results or genetic variants that cancel out and are dismissed as unimportant. So until now, it has been easier to exclude people with multiple ancestry from GWAS to avoid being confused by different patterns of genetic variants.

The tractor, however, allows researchers to report ancestry accurately so that mixed individuals can be included in large-scale gene discovery efforts. The software colors pieces of each person’s chromosomes based on their ancestral origin, which researchers can derive from reference genome sequences, and uses that information in a new GWAS model. “The tractor takes into account the backbone of the ancestry of each genetic variant so that we can properly calibrate the GWAS results to find causal variants in specific population groups,” Atkinson said.

Tractor also provides lineage-specific effect size estimates, which is not possible in a standard GWAS. “Instead of getting a weighted average of the disease risk effect size for a particular genetic variant, Tractor can determine the magnitude or small size of a variant’s effect in various ancestry groups. “said Atkinson. “It will be instructive for establishing genetic risk scores in various populations.”

Another advantage of Tractor is its ability to enhance the potency of GWAS by detecting risky gene variants in multiple ancestries. “With Tractor, we can achieve stronger disease association signals by taking advantage of ancestral genomic differences,” Atkinson added.

“Tractor advances existing methodologies to study the genetics of complex disorders in diverse and minority populations,” she said. “We hope this method will increase the inclusion of mixed participants in large-scale association studies in the future.”

Reference: Atkinson EG, Maihofer AX, Kanai M, et al. The tractor uses local ancestry to allow inclusion of mixed individuals in GWAS and to increase horsepower. Nat Genet. 2021. doi: 10.1038 / s41588-020-00766-y.

This article was republished from the following materials. Note: The material may have been modified for its length and content. For more information, please contact the cited source.


Comments are closed.