Missing Heritability and Polygenic Architecture

When the first genome-wide association studies tallied the trait variance explained by their genome-wide significant variants, the total fell far short of the heritability estimated from family and twin studies - a gap that became known as the 'missing heritability' problem. Resolving it reshaped how researchers think about the genetic architecture of common traits, pointing toward a highly polygenic model in which very many variants each contribute a tiny effect.

Definition

Missing heritability is the discrepancy between the heritability of a trait estimated from family or twin studies and the smaller proportion explained by variants individually reaching genome-wide significance; polygenic architecture is the underlying model in which a trait is influenced by very large numbers of variants, most of small effect.

Scope

This topic covers what missing heritability means, the candidate explanations (many undetected small-effect common variants, rare variants, structural variation, gene-gene and gene-environment interaction, and overestimated heritability), and the methods - such as genome-wide complex trait analysis - that showed much of the gap reflects common variants below the significance threshold. It is a conceptual and methodological reference, not clinical guidance.

Core questions

• Why did genome-wide significant variants explain only a fraction of family-based heritability?
• How much of the gap is hidden in common variants too small to reach significance?
• What roles might rare variants, structural variation, or interactions play?
• Could family-based heritability estimates themselves be inflated?
• What does the resolution imply for the genetic architecture of complex traits?

Key concepts

• Narrow-sense heritability
• Family- and twin-based heritability estimates
• SNP-based heritability
• Genome-wide complex trait analysis (GCTA / GREML)
• Common variants below the significance threshold
• Rare and structural variation
• Gene-gene and gene-environment interaction

Key theories

Polygenic (infinitesimal) architecture of complex traits

Common traits are influenced by a very large number of variants, the great majority with effects too small to individually exceed genome-wide significance, so methods that aggregate genome-wide variation recover far more heritability than counting top hits alone; this reframed missing heritability as largely hidden in undetected small-effect common variation.

Mechanisms

Several non-exclusive mechanisms were proposed for the gap. Most consequentially, a trait can be highly polygenic, with thousands of common variants each of tiny effect that fall below the genome-wide significance threshold in any finite sample; methods that estimate variance captured by all genotyped variants jointly - rather than only significant ones - showed for traits such as height that common SNPs collectively account for a large share of heritability. Other contributors include rare variants poorly tagged by genotyping arrays, structural variants such as copy-number changes, interactions between variants or between genes and environment, and the possibility that family-based estimates are themselves inflated by shared environment or non-additive effects. As samples grew into the hundreds of thousands and beyond, more loci crossed significance and the explained fraction rose, consistent with the polygenic interpretation.

Clinical relevance

The polygenic view that emerged from this debate underlies how aggregate genetic effects - for example in polygenic scores - are conceptualised and interpreted in research. This topic describes genetic architecture and is not a basis for individual risk prediction or clinical decision-making.

Evidence & guidelines

Understanding here rests on methodological reviews and primary analyses rather than clinical guidelines. Manolio et al. (2009) framed the problem and catalogued candidate explanations; Yang et al. (2010) demonstrated with genome-wide complex trait analysis that common SNPs explain much of height's heritability; and Visscher et al. (2012, 2017) synthesised how growing sample sizes and polygenic methods progressively narrowed the gap.

History

The phrase 'missing heritability' was popularised by a 2009 review after early GWAS for common diseases and traits left most family-based heritability unexplained. A turning point came in 2010 when genome-wide complex trait analysis showed that common variants, considered jointly, captured far more of height's heritability than the significant hits alone, recasting much of the 'missing' fraction as merely 'hidden' below the significance threshold. Subsequent biobank-scale studies confirmed the highly polygenic architecture of common traits while leaving residual debate about the contribution of rare and structural variation.

Debates

Is the remaining gap due to rare variants or to still-undetected common variants?

After SNP-based heritability accounted for much of the shortfall, debate continued over whether the residual reflects rare variants poorly captured by arrays, structural variation and interactions, or simply common variants awaiting larger samples - distinctions with different implications for study design.

Key figures

• Teri Manolio
• Peter Visscher
• Jian Yang
• Naomi Wray
• David Goldstein

Seminal works

• manolio-2009
• yang-2010
• visscher-2012

Frequently asked questions

What does 'missing heritability' actually mean?

It is the gap between how heritable a trait appears from family and twin studies and the much smaller share explained by the individual variants that reached genome-wide significance in early GWAS.

Was the heritability really missing?

Largely no - methods that aggregate all common variants showed much of it was hidden in many small-effect variants below the significance threshold, though rare and structural variation may account for some of the remainder.