Biology Atlas

Map / Open register

question · deep · partly written

Which chromatin change explains a GWAS signal?

eQTLs colocalise with a minority of GWAS signals, chromatin-accessibility QTLs add some loci mostly without a transcript link, and the numbers use different denominators.

The short version

Reported shares of GWAS signals with a known mechanism range from about 8% to a little over half, depending on assay and denominator. The numbers cannot be put on one scale. 1234

Reported, not yet replicated

The figures and their denominators more

Mostafavi and colleagues quote a GTEx analysis: only 43% of GWAS hits (median 21% per trait) colocalised with eQTLs, and cell type-specific eQTLs colocalised with about 8% of hits. Mu and colleagues, in immune cells from 48 individuals, report 56.8% of 8,271 GWAS loci colocalising with caQTLs, eQTLs or both. Currin and colleagues, in liver, report 844 colocalised caQTL signals and estimate 673 of 4,330 (16%) LD-clumped GWAS signals colocalised with caQTLs. Gschwind and colleagues (preprint) report 16% (range 10 to 22%) of fine-mapped eQTL variants linked to the correct gene by their model and 61% precision against 52% for the second best on blood-trait credible sets. 1234

Reported, not yet replicated

Reading across them deep

The first is a share of GWAS hits explained by eQTLs, the second a share of loci with a colocalised QTL of either type, the third a share of clumped signals, the fourth a share of eQTL variants correctly linked to a gene. They answer different questions. The immune caQTL abstract says most GWAS-colocalising caQTLs lacked eQTL support, limiting causal inference from chromatin data alone. 1234

Reported, not yet replicated

Observation

In one analysis by the GTEx Consortium only 43% of GWAS hits (median 21% per trait) colocalised with eQTLs, and cell type-specific eQTLs colocalised with about 8%; these figures are reported second-hand through a 2023 paper. In immune cells, 56.8% (4,696 of 8,271) of GWAS loci colocalised with caQTLs, eQTLs or both, with 2,015 caQTL-only, 1,149 eQTL-only and 1,532 both. In liver, 844 caQTL signals colocalised with GWAS signals, of which 231 also with an eQTL. A model linking enhancers to genes identified on average 16% of fine-mapped eQTL variants with the correct gene (preprint). 1234

Reported, not yet replicated

Hypotheses

eQTLs miss signals because of context and selection

Trait-relevant variants act in specific cell states or are depleted among detectable eQTLs through selection, so eQTL maps are biased toward different variants. 1

Predicts. eQTL hits cluster near transcription start sites while GWAS hits do not.

Would falsify. Deep sampling of cell types and contexts closing the gap.

Reported, not yet replicated

Chromatin accessibility explains additional loci

caQTLs colocalise with more GWAS loci than eQTLs, nominating causal genes for some unexplained loci. 23

Predicts. More colocalisations with caQTLs than eQTLs in matched tissue.

Would falsify. caQTL-GWAS loci that fail validation of regulatory activity at several loci.

Reported, not yet replicated

Accessibility alone limits causal inference

Most GWAS-colocalising caQTLs lack eQTL support, so chromatin data alone cannot show the gene. 2

Predicts. Colocalised caQTLs without eQTLs remain unexplained at the transcript level.

Would falsify. Direct transcript readouts that connect most such loci.

Reported, not yet replicated

What is missing

A common denominator

The 43%, 21%, 8%, 56.8%, 16% (liver clumped signals) and 16% (eQTL variants linked to the right gene) figures use different GWAS sets, cell types, colocalisation methods and units; no opened study states the fraction of GWAS variants that act through accessibility to change a transcript. 1234

Reported, not yet replicated

End-to-end validation

Validation by reporter or CRISPRi was done at a handful of loci in the liver study. 3

Reported, not yet replicated

The GTEx paper

The GTEx figures were read second-hand and the GTEx paper was not opened here. 1

Reported, not yet replicated

What is known

eQTLs explain only a small fraction of GWAS signals; trait-associated variants are enriched in tissue-specific epigenomic marks; caQTLs add loci; most of the added loci lack an eQTL link. 1523

Reported, not yet replicated

What is not

The share of GWAS signals mediated by a chromatin change that alters a transcript, per cell state, and whether current enhancer-gene linking models are accurate beyond a minority. 124

Reported, not yet replicated

What would settle it

Matched, multi-assay data (accessibility, nascent transcription and perturbation) in disease-relevant cell states, with a stated denominator and validation at many loci. 23

Our inference

Still open here

Not yet written: Who works on it.

Sources

  1. Mostafavi, H., Spence, J.P., Naqvi, S., Pritchard, J.K.. Systematic differences in discovery of genetic effects on gene expression and complex traits. Nature Genetics (2023). paper; opened 2026-11-17 all records citing this
  2. Mu, Z., Randolph, H.E., Aguirre-Gamboa, R., et al., Li, Y.I.. Impact of disease-associated chromatin accessibility QTLs across immune cell types and contexts. Cell Genomics (2026). paper; opened 2026-11-17 all records citing this
  3. Currin, K.W., Perrin, H.J., Pandey, G.K., et al., Mohlke, K.L.. Genetic effects on chromatin accessibility uncover mechanisms of liver gene regulation and quantitative traits. Genome Research (2025). paper; opened 2026-11-17 all records citing this
  4. Gschwind, A.R., Mualim, K.S., Karbalayghareh, A., et al., Engreitz, J.M.. An encyclopedia of enhancer-gene regulatory interactions in the human genome. bioRxiv (2023). preprint, not peer reviewed; opened 2026-11-17 all records citing this
  5. Kundaje, A., Meuleman, W., et al., Kellis, M.. Integrative analysis of 111 reference human epigenomes. Nature (2015). paper; opened 2026-11-17 all records citing this
Where it sits on the map
History of this record (1)

This record has been changed 1 time, on the dates below. The note kept against each change is written for the people who maintain the atlas, in their own shorthand, so it is not shown here; it is in the record itself.

Kind
big-open
Register
open-question
Status
partly written
Last updated
2026-11-17
Record
content/questions/which-chromatin-change-explains-a-gwas-signal.json