Genetic architecture: the shape of the genetic contribution to human traits and disease

genetic architecture
complex traits
GWAS
heritability
rare variants
pleiotropy
gene-environment interaction
  • This review defines genetic architecture by four components: the number of causal variants (polygenicity), the distribution of their effect sizes, their allele frequency spectrum, and the types of genetic and environmental interactions (dominance, epistasis, GxE).
  • It highlights that complex traits are highly polygenic and influenced by variants across the entire frequency spectrum, addressing “missing heritability” by pointing to the role of rare variants, non-additive effects, and Gene-by-Environment (GxE) interactions.
  • The authors emphasize that pleiotropy (one variant affecting multiple traits) is widespread among common variants, discussing how techniques like Mendelian Randomization (MR) are essential for distinguishing causation from pleiotropy in the complex genetic landscape.
Published

23 January 2026

PubMed: 29225335
DOI: 10.1038/nrg.2017.101
Overview generated by: Gemini 2.5 Flash, 26/11/2025

Key Findings

This comprehensive review synthesizes the field’s understanding of genetic architecture—the characteristics of genetic variation responsible for heritable phenotypic variability—in the era following the advent of large-scale Genome-Wide Association Studies (GWAS) and Next-Generation Sequencing (NGS). The authors systematically define the components of genetic architecture and explain how recent technological advances have begun to reveal the complex interplay of factors contributing to human traits and diseases.

Core Components of Genetic Architecture

The genetic architecture of any complex trait is defined by four interacting components, which the review explores in depth:

  1. Number of Causal Variants: The sheer count of genetic variants (SNPs, indels, SVs) that collectively affect the trait. Most complex traits are highly polygenic, involving thousands of variants.
  2. Effect Size Distribution: The magnitude of the effect that each variant contributes to the phenotype. GWAS has revealed that many common variants have small, additive effects, while rare variants often have large effects.
  3. Allele Frequency Spectrum: The distribution of causal variant frequencies in the population. Common traits are often influenced by variants across the entire frequency spectrum, supporting the ‘common disease, common variant’ and ‘common disease, rare variant’ hypotheses in tandem.
  4. Interactions: The complexity added by non-additive relationships:
    • Allelic Interactions: Dominance (interaction between alleles at the same locus).
    • Locus Interactions: Epistasis (interaction between alleles at different loci).
    • Environmental Interactions: Gene-by-Environment (GxE) interaction.

Impact of Modern Genomic Technologies on Architecture Discovery

The review highlights how different technologies have been instrumental in characterizing specific aspects of the genetic architecture:

GWAS and Common Variant Architecture

  • GWAS Success: GWAS has been highly successful in identifying thousands of common, low-effect variants for hundreds of traits, confirming the extreme polygenicity of complex traits.
  • Missing Heritability: The review addresses the historical problem of “missing heritability”—the gap between heritability estimated from twin/family studies (broad-sense heritability) and that explained by all detected common SNPs (SNP-heritability). Explanations include:
    • The contribution of rare variants missed by GWAS arrays.
    • The residual influence of non-additive effects (dominance and epistasis) captured by family studies but not fully by linear GWAS models.
    • The contribution of structural variation and gene-environment interactions.
  • Locus Heterogeneity: GWAS often reveals multiple independent associated signals within the same locus, indicating allelic series or complex local regulation.

Next-Generation Sequencing (NGS) and Rare Variants

  • Sequencing Role: NGS studies (e.g., whole-exome sequencing, whole-genome sequencing) are essential for characterizing the role of rare variants.
  • Burden Tests: These tests aggregate the effects of multiple rare variants within a single gene or region. The review notes that rare, high-penetrance variants often reside in genes under strong negative selection, explaining why their overall contribution to population variance (though individually large) may be limited.
  • Clinical Relevance: Rare variants are crucial for understanding Mendelian disease and for identifying genes with large effects that are strong candidates for drug targets.

Complexity of Genetic Effects

Pleiotropy and Shared Genetic Etiology

  • Widespread Pleiotropy: The authors emphasize that pleiotropy (a single genetic variant affecting multiple distinct traits) is the norm, not the exception, for common variants identified by GWAS. This is supported by analyses showing extensive genetic correlation between traits.
  • Confounding: Pleiotropy complicates causal inference. The review discusses how techniques like Mendelian Randomization (MR) are used to distinguish true causal effects from horizontal pleiotropy (a single variant affecting multiple outcomes through different pathways).

Gene-Environment (GxE) Interactions

  • Definition: GxE occurs when the effect of a genetic variant on a phenotype depends on the individual’s environment (e.g., diet, smoking, stress).
  • Detection Challenge: GxE interactions are notoriously difficult to detect and estimate accurately due to requiring large samples with precise environmental measures. The review notes that population-based cohorts like the UK Biobank are vital for making progress in this area.

Future Directions and Clinical Goals

The review concludes by outlining the necessary steps to fully characterize genetic architecture and achieve the field’s clinical goals:

  1. Comprehensive Mapping: Moving from association studies to causal variant identification, focusing on non-coding variants and improving fine-mapping methods.
  2. Accounting for Non-Additivity: Developing statistical methods that are better powered to detect and estimate dominance and epistatic effects in large cohorts.
  3. Integrating Environment: Robustly incorporating environmental exposure data into models to quantify the contribution of GxE interactions and improve personalized risk prediction.
  4. Clinical Translation: Leveraging the understanding of genetic architecture to improve disease screening, diagnosis, prognosis, and therapeutic development. This includes prioritizing genes for drug development based on the magnitude and specificity of their genetic effects.