Statistical & Population Genetics Visualizer
Genome-Wide Association Studies (GWAS) & Statistical Genetics
Interactive exploration of genome-wide association scans, single-variant OLS regression, population stratification PCA correction, Linkage Disequilibrium (LD) fine-mapping, and Polygenic Risk Scores (PRS).
Type 2 Diabetes (T2D) (Type 2 Diabetes Mellitus)
Classic metabolic polygenic disease. Lead signal at rs7903146 in TCF7L2 impairs pancreatic β-cell insulin exocytosis and GLP-1 expression. TCF7L2 rs7903146-T confers a 1.37-fold increased risk per allele (p = 1.2 × 10⁻¹⁹), representing the strongest single common risk locus identified for T2D.
| Genomic Locus | chr10:114,758,349 |
|---|---|
| Alleles | T / C |
| Minor Allele Freq (MAF) | 30.0% |
| Effect Size (β) | +0.340 (SE ±0.018) |
| Association P-Value | 1.20e-19 |
| Nearest Gene | TCF7L2 |
| Consequence | intron variant (enhancer) |
For quantitative traits, single-SNP association solves the additive linear model:
The Wald test statistic and two-tailed p-value evaluate the null hypothesis :
Without correcting for continental ancestry or cryptic relatedness, systematic allele frequency divergence creates massive genome-wide false-positive associations ().
Including the top 10–20 Principal Components () or using Linear Mixed Models () projects out ancestry confounding, recovering a pristine null distribution ().
Individual Polygenic Risk Scores (PRS) sum the effect sizes of genome-wide risk variants:
Individuals in the top 5% risk percentile exhibit 3–5× higher lifetime disease probability, comparable to rare high-penetrance monogenic mutations (e.g. LDLR or BRCA1).
1. The Bonferroni Multiple Testing Threshold
Testing ~1,000,000 independent Linkage Disequilibrium (LD) blocks across the human genome at family-wise error rate yields the canonical genome-wide significance threshold: ().
2. Genomic Inflation & Confounding ()
The Genomic Inflation Factor measures systematic test statistic deviation. Elevated indicates uncorrected population stratification, batch artifacts, or extreme polygenicity (analyzed via LD Score Regression).
3. Linkage Disequilibrium (LD) & Fine-Mapping
Because adjacent variants are co-inherited in LD blocks (), GWAS identifies genomic loci rather than causal variants. Statistical fine-mapping (e.g. SuSiE, PAINTOR) and functional epigenomics (ChIP-seq, eQTLs, OpenSpliceAI) pinpoint true causal mechanisms.