Sign In

Communications of the ACM

ACM TechNews

Making Genetic Prediction Models More Inclusive

View as: Print Mobile App Share:
People from a variety of backgrounds.

The new model that takes into account genetic information from people of a wider variety of genetic backgrounds.

Credit: iStock

Computer scientists at the Massachusetts Institute of Technology (MIT) created a more inclusive and more accurate genetic prediction model by using computational and statistical techniques to study individuals' unique genetic profiles.

The model used UK Biobank genetic data on more than 280,000 people and another dataset of around 81,000 held-out individuals, and was evaluated across 60 traits.

This allowed the researchers to include people of admixed ancestry, who accounted for close to 10% of the UK Biobank dataset.

The model improved predictions by an average of 61% for people of African ancestry, 18% for people of admixed ancestry, 11% for people of South Asian ancestry, and 5% for white British individuals.

Said MIT's Yosuke Tanigawa, "When you bring all the individuals together in the training set, everybody contributes to the training of the polygenic score modeling on equal footing."

From MIT News
View Full Article


Abstracts Copyright © 2023 SmithBucklin, Washington, D.C., USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account