Genome-Wide Association Studies: How Science Links Genes to Disease

Genome-wide association studies (GWAS) have transformed our understanding of the genetic architecture of common diseases and traits. By scanning hundreds of thousands to millions of genetic variants across the entire genome in large populations, GWAS have identified thousands of genetic risk loci for conditions ranging from type 2 diabetes to depression to height.

How GWAS Works

The basic design of a GWAS involves genotyping (determining the specific variant at each of hundreds of thousands to millions of genetic positions) in a large group of individuals with the disease of interest (cases) and a comparison group without the disease (controls). Statistical tests determine which variants occur at significantly different frequencies between cases and controls.

Modern GWAS use microarray platforms that genotype 500,000-1,000,000+ SNPs per sample, combined with statistical imputation to infer untyped variants based on linkage disequilibrium patterns. The resulting statistical threshold for genome-wide significance is p less than 5x10^-8, accounting for the massive multiple comparisons problem.

Key GWAS Discoveries

GWAS have identified thousands of loci associated with common diseases. For type 2 diabetes, more than 500 associated loci have been identified, collectively explaining approximately 20% of disease heritability. For Alzheimer's disease, GWAS confirmed APOE4 as the strongest genetic risk factor and identified dozens of additional loci implicating complement, lipid metabolism, and immune pathways. For inflammatory bowel disease, over 200 loci have been identified, revealing unexpected involvement of innate immunity genes. These discoveries have generated novel biological insights and drug targets.

Polygenic Risk Scores

GWAS findings can be aggregated into polygenic risk scores (PRS)—weighted sums of risk alleles across many loci—that capture an individual's inherited predisposition to a disease. PRS are increasingly being incorporated into clinical medicine for cardiovascular disease risk stratification and cancer screening decision-making.

Limitations of GWAS

GWAS have important limitations. Most GWAS have been conducted predominantly in European populations, limiting generalizability. Most common disease-associated variants have small individual effects. GWAS identifies association, not causation. And most associated variants are in non-coding regions, making functional interpretation challenging.

For more on genomics research and its clinical applications, visit our blog.