Skip to content

Tissue-Partitioned Heritability

Summary

Tissue-partitioned heritability methods take GWAS summary statistics and ask not which gene but which tissues and cell types a trait's genome-wide genetic signal is concentrated in. Stratified LD score regression (S-LDSC) partitions SNP-heritability across overlapping functional annotations, and its LDSC-SEG extension builds annotations from tissue- and cell-type-specifically expressed genes to nominate disease-relevant tissues. Together they convert a diffuse polygenic signal into a ranked list of candidate tissues of action, complementing single-locus methods that work one variant at a time.

What It Does

  • Stratified LD score regression (S-LDSC) partitions a trait's SNP-heritability across many (overlapping) functional annotation categories — coding, conserved, enhancer/promoter, cell-type-specific regulatory marks — using only GWAS summary statistics plus an ancestry-matched LD reference. The output per category is an enrichment: the share of heritability it captures relative to the share of SNPs it contains.
  • LDSC-SEG (specifically expressed genes) extends this to tissue/cell-type identification. For each tissue or cell type it builds an annotation from the genes most specifically expressed there, then tests which annotation is most enriched for the trait's heritability. The tissues whose specifically-expressed-gene regulatory regions are most heritability-enriched are nominated as trait-relevant.

The defining feature is that it uses the whole polygenic signal — every SNP, not just genome-wide-significant hits — so it can localise traits whose individual loci are each too weak to interpret.

Why It Matters (the variant → tissue rung)

Single-locus methods answer "which gene / which variant." Tissue-partitioned heritability answers a different, aggregate question: "which tissue or cell type does inherited risk act through?" — before any experimental perturbation. It is a genetics-first result (a GWAS is the input; a tissue is the output), which is why it belongs alongside colocalization and QTL mapping rather than with expression-atlas characterisation.

For cardiovascular GWAS, functional and expression-based partitioning typically implicates vascular/arterial, adipose, hepatic, and immune regulatory programs , tying inherited CVD risk back to the same tissues studied through lipids and cardiac fat.

Inputs and Requirements

  • GWAS summary statistics (no individual-level genotypes required).
  • LD scores from an ancestry-matched reference panel (e.g. 1000 Genomes); mismatched LD references bias the estimates.
  • Annotations: a baseline functional model plus the tissue-/cell-type-specific annotations (e.g. specifically-expressed-gene sets, or chromatin/enhancer maps).

Relation to Other Vault Methods

Caveats

  • Identifies enriched categories, not causal genes — a tissue nomination is a hypothesis about where to look, not a mechanism.
  • Correlated tissues (e.g. related immune cell types) are hard to disentangle, and annotations overlap.
  • Tissue-enrichment narratives are easy to over-read: believable when the enrichment is specific and replicates, weak when a story is fit post hoc (see the "just-so tissue-enrichment stories" caution on Polygenic Subtyping of Cardiovascular Disease).

See Also

Citations

[1] Finucane, H. K., et al. (2015). Partitioning heritability by functional annotation using genome-wide association summary statistics. Nature Genetics, 47, 1228–1235. DOI: 10.1038/ng.3404. Supports: the S-LDSC method and the definition of heritability enrichment above. Location: canonical primary source for S-LDSC — bibliographic details recorded from established method knowledge and not re-inspected in this session; verify before relying on specific figures.

[2] Finucane, H. K., et al. (2018). Heritability enrichment of specifically expressed genes identifies disease-relevant tissues and cell types. Nature Genetics, 50, 621–629. DOI: 10.1038/s41588-018-0081-4. Supports: the LDSC-SEG specifically-expressed-genes extension and tissue/cell-type nomination above. Location: canonical primary source for LDSC-SEG — recorded from established method knowledge and not re-inspected in this session; verify before relying on specific figures.