GWAS Meta-Analysis of Admixed Populations (GMAX) uses local ancestry inference to identify associated loci in GSCAN meta-analysis

Research Poster Health & Life Sciences 2025 Graduate Exhibition

Presentation by Natashia Benjamin

Exhibition Number 195

Abstract

People with mixed ancestry, known as admixed populations, inherit their genes from multiple continents, creating a unique genetic mosaic. However, most genetic studies focus on European populations, making it harder to identify important genetic factors in diverse groups. Existing methods like TRACTOR demonstrates that incorporating local ancestry information in genome wide association studies improves power of discovering variant-trait associations, especially in admixed populations. Despite this fact, no method currently incorporates local ancestry information in meta-analysis. To address this, we developed a method, GMAX Local Ancestry Inference (GMAX-LAI), to integrate local ancestry across admixed genomes for GWAS meta-analysis. Our method estimates ancestry proportions at a given variant for admixed study by decomposing allele frequencies as a weighted sum of continental ancestry frequencies. Compared to RFMix, a widely used individual level LAI method, our approach provides comparable estimates. These estimates are then incorporated into a mixed-effect meta-regression model, capturing genetic effects in our meta-analysis. We applied our method to GSCAN (GWAS & Sequencing Consortium of Alcohol and Nicotine use) traits in diverse populations (55% European, 15% African American, 6% Latino/Hispanic American, 24% East Asian). We observe significant ancestry proportion difference across studies revealing substantial study-specific local ancestry genetic structure. By meta-analyzing 121 studies, we identified 1,560 loci, including novel associations at MSANTD4, TENM4, GAS2L3, STARD9, UXS1 and NLGN1, which were missed by both the fixed-effect model and the global ancestry-based model, MEMO. Our findings highlight the value of local ancestry integration in GWAS meta-analysis, improving discovery in admixed populations.

Importance

Genome wide association studies (GWAS) are used to identify genomic variants that are statistically associated with a particular disease or trait. Over 75% of GWAS participants are of European descent, leading to European-biased genetic models. However, these models perform sub-optimally in non-European populations, limiting their accuracy and utility. While local ancestry information has been shown to improve analyses within admixed populations, existing methods require individual-level data and cannot be applied to GWAS summary statistics. This gap means that no current GWAS meta-analysis methods incorporate local ancestry, restricting discovery in diverse populations. Addressing this limitation is essential for more equitable and accurate genetic research, improving disease risk prediction and personalized medicine for underrepresented populations.

DEI Statement

Given the ancestral disproportionality in biobank data, a great majority of genetic studies are conducted mainly on the European population. Our work directly promotes diversity, equity, and inclusion (DEI) by developing methods that better analyze genetic data from admixed and underrepresented groups. By incorporating local ancestry information, we improve genetic discovery, disease risk prediction, and personalized medicine for populations often overlooked in genomic studies. This research ensures that biomedical advancements benefit all groups equitably, rather than reinforcing health disparities. By addressing gaps in genetic diversity and representation, our work aligns with DEI principles, fostering more inclusive, accurate, and impactful genetic research that serves global populations.