Amid the waves of technological and scientific advancements, the significance of genetic variation remains paramount. About 0.1% of our genetic diversity plays a crucial role in determining susceptibility to certain diseases. Despite this, most genetic models used to predict polygenic scores are tailored around European ancestry data, often marginalizing individuals of non-European or admixed backgrounds.
MIT’s groundbreaking research seeks to bridge this gap. By integrating genetic information from a broader range of ancestral backgrounds, their innovative model showcases enhanced accuracy, especially for populations usually sidelined in genetic studies. According to Prof. Manolis Kellis from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), the new model exhibited a striking 60% accuracy surge for African ancestry and approximately 18% for those of mixed genetic heritage.
Building upon the legacy of the Human Genome Project and the subsequent cohort-based studies, Kellis underscores the inherent limitations of most genome-wide association studies. Given that many of these studies majorly comprised European subjects, their findings are often not applicable to non-European populations, rendering the genetic risk predictions skewed.
Kellis points out the fundamental flaw: “If you are an individual of African descent, of Latin American descent, of Asian descent, then you are currently being left out by the system.” This systemic exclusion leads to preventable hardships and health risks.
Although some contemporary research has initiated the creation of ancestry-specific models, they still fall short in accounting for those with mixed genetic backgrounds. MIT’s revolutionary approach thrives on its individual-centricity, making arbitrary groupings obsolete. Leveraging computational techniques, the research seamlessly integrated almost 10% of the admixed individuals from the UK Biobank dataset, which formed the foundation of the study.
Armed with data from over 280,000 individuals collected by the UK Biobank, MIT’s new model’s prowess was tested across 60 genetically-based traits. The results were astoundingly positive, most notably for those of African ancestry, reflecting a 61% improvement.
The promising strides made by MIT herald a new era of more inclusive and effective genetics-based health predictions. Combining these findings with traditional risk factors can potentially revolutionize healthcare, guiding individuals in managing and preempting diseases more efficiently.
Yosuke Tanigawa, an MIT postdoc, emphasized, “Our work highlights the power of diversity, equity, and inclusion efforts in the context of genomics research.” The team is now gearing up to delve deeper, incorporating even more data to further the cause of inclusive genetic research.
With robust funding from the National Institutes of Health, MIT’s endeavor is just the beginning of a more inclusive chapter in human genetics research.