The GTEx Consortium atlas of genetic regulatory effects across human tissues
François AguetiD, Alvaro BarbeiraiD, Rodrigo BonazzolaiD, Andrew Anand BrowniD, Stephane E. CasteliD, Brian JoiD, Silva KaselaiD, Sarah Kim-HellmuthiD, Yanyu LiangiD, Meritxell OlivaiD, Princy ParsanaiD, Elise D. FlynniD, Laure FrésardiD, Eric R. GamazoniD, Andrew R. HameliD, Yuan HeiD, Farhad HormozdiariiD, Pejman MohammadiiD, Manuel Muñoz-AguirreiD, YoSon ParkiD, Ashis SahaiD, Ayellet V Segrć, Benjamin J. StroberiD, Xiaoquan WeniD, Valentin WucheriD, , Tuuli LappalaineniD
Abstract
The Genotype-Tissue Expression (GTEx) project was established to characterize genetic effects on the transcriptome across human tissues, and to link these regulatory mechanisms to trait and disease associations. Here, we present analyses of the v8 data, based on 17,382 RNA-sequencing samples from 54 tissues of 948 post-mortem donors. We comprehensively characterize genetic associations for gene expression and splicing in cis and trans, showing that regulatory associations are found for almost all genes, and describe the underlying molecular mechanisms and their contribution to allelic heterogeneity and pleiotropy of complex traits. Leveraging the large diversity of tissues, we provide insights into the tissue-specificity of genetic effects, and show that cell type composition is a key factor in understanding gene regulatory mechanisms in human tissues.
02 /Specification & Classification
Click any tag to browse matching papers- Citations
- 366
- Organism
- Assay
- Research Focus
- Approach
In-depth summary
Claude Opus 5.5 · max effort · 8 Nov 2026 · from the full text of version 1 (3 Oct 2019) · CC BY 4.0
GTEx v8 maps expression and splicing QTLs in 49 tissues from 838 donors: cis-eQTLs cover 94.7% of protein-coding genes, and cross-tissue QTL sharing mirrors cell-type composition.
Complex-trait risk is driven mainly by non-coding loci of largely uncharacterized regulatory function, and earlier large-scale QTL maps centered on whole blood and blood cell types. Extending the GTEx pilot and mid-stage releases, v8 provides a multi-tissue catalog of cis and trans genetic effects on expression and splicing, with fine-mapping, cell type interaction QTLs, and GWAS colocalization linking variants to target genes, tissues, and mechanisms.
Key results
ShownGenetic regulation of expression is nearly universal: almost every protein-coding gene has a cis-eQTL in at least one tissue, many genes carry several independent cis-eQTLs, and discovery still scales with tissue sample size.
23,268 cis-eGenes at 5% FDR (18,262 protein coding, 94.7%; 5,006 lincRNA, 67.3% of detected) and 4,278,636 cis-eVariants; cis-sQTLs for 14,424 genes; discovery vs sample size Spearman's rho = 0.95 (eQTL), 0.92 (sQTL); in the best-powered tissues up to 50% of eGenes have multiple independent cis-eQTLs. · QTL discovery
ShownRegulatory QTLs are enriched among complex-trait associations, most strongly for likely causal fine-mapped variants, and colocalization nominates candidate target genes and mechanisms for thousands of GWAS signals.
GWAS catalog overlap 1.46-fold (cis-eQTL), 1.86-fold (cis-sQTL), 6.97-fold (trans-eQTL); QTLEnrich 1.43-fold (SE=0.04, eQTL) and 1.52-fold (SE=0.04, sQTL); S-LDSC with continuous fine-mapping annotation 11.1-fold (SE=1.2, eQTL) and 14.2-fold (SE=2.4, sQTL); of 5,385 GWAS associations over 87 traits, 43% colocalized (enloc) with a cis-eQTL, 23% with a cis-sQTL. · Genetic regulatory effects mediate complex trait…
ShownMulti-method statistical fine-mapping narrows many cis-eQTLs to a handful of candidate causal variants, and confidently fine-mapped variants are more enriched among experimentally validated regulatory variants than lead eVariants.
dap-g 90% credible sets contain a median of 6 variants, and 9.3% of cis-eQTLs have a variant with posterior >0.8; 24,740 cis-eQTLs (7,709 unique variants) exceed posterior 0.8 under all of CaVEMaN, CAVIAR and dap-g; this consensus set had the highest MPRA/SuRE enrichment, with overlapping CIs. · Fine-mapping
ShownQTL sharing across tissues parallels cell type composition similarity and is predictable from annotation and QTL features, while genotype-by-cell-type interaction mapping finds interacting eQTLs (ieQTLs), including eGenes and GWAS colocalizations missed in bulk tissue.
1120 neutrophil ieQTLs (whole blood) and 1087 epithelial ieQTLs (transverse colon), 76 and 229 involving eGenes without a bulk QTL; 111 GWAS loci (about 32%) colocalized only with neutrophil ieQTLs, not whole-blood cis-eQTLs; cis-QTL activity in a query tissue predicted with median AUC 0.779 (eQTL) and 0.807 (sQTL). · Tissue-specificity of genetic regulatory effects; From tissues to cell types
ShownTrans-eQTLs are scarce and frequently tissue-restricted, but many arise through local cis-eQTL effects, and their enrichment in CTCF binding sites suggests chromatin-mediated distal regulation.
143 trans-eGenes (121 protein coding, 22 lincRNA) and 29 trans-sGenes at 5% FDR; 47 of 49 testis trans-eGenes were absent from all other tissues even at 50% FDR; lead trans-eVariants 5.9x enriched for being cis-eVariants (Fisher's p = 5.03×10^−22); mediation implies 77% of lead trans-eVariants doubling as cis-eVariants act via the cis-eQTL. · QTL discovery; Functional mechanisms of QTL associations
Methods, data and reproducibility
Bulk RNA-seq of 49 tissues (≥70 donors each) and whole-genome sequencing genotypes were used to map cis/trans expression and LeafCutter-based splicing QTLs, with allelic expression for validation. Follow-up used three fine-mapping methods, QTLEnrich and S-LDSC, enloc/COLOC colocalization, PrediXcan, mashr sharing, and xCell-based cell type interaction QTLs.
- GTEx v8 bulk RNA-seq of post-mortem donor tissues: 52 tissues (including 11 brain regions) plus two cell lines · human · n = 17,382 samples from 948 donors · dbGaP phs000424.v8
- Donors with WGS genotypes and RNA-seq (85.3% European American, 12.3% African American, 1.4% Asian American); QTLs mapped in 49 tissues sampled in ≥70 donors · human · n = 838 donors; 15,253 samples (15,201 in QTL tissues) · dbGaP phs000424.v8
- WGS variant calls after QC and phasing, sequenced to median 32x depth · human · n = 43,066,422 SNPs and 3,459,870 small indels · dbGaP phs000424.v8
- Public GWAS summary statistics for binary and continuous complex traits, plus GWAS catalog variants for overlap enrichment · human · n = 87 GWAS
- Functional annotations: ENCODE TADs and transcription factor ChIP-seq, Ensembl Regulatory Build features, and chromatin states
- MPRA and SuRE experimentally validated causal regulatory variants used to benchmark fine-mapping
- External eQTL data from purified neutrophils (used to replicate neutrophil ieQTLs) and a separate set of five sorted blood cell types (used for the allelic heterogeneity analysis) · n = 5 sorted cell types (second dataset)
- UK Biobank rare-variant trait associations compared against cis-eQTL colocalization · human (UK Biobank)
- Methods & tools
- LeafCutter, CaVEMaN, CAVIAR, dap-g, mashr, QTLEnrich, Stratified LD score regression (S-LDSC), enloc, PrediXcan, COLOC, xCell, Haplotype-level allelic expression (AE) analysis
- Data & accessions
- http://www.gtexportal.org/https://gtexportal.org/home/protectedDataAccesshttps://www.gtexportal.org/home/samplesPage
- Reagents & environment
- Biological material: post-mortem human donor tissues across 54 tissue types, including 11 brain regions and 2 cell lines; no software versions or containers are stated.
- Experimental conditions
- Sequencing parameters only: WGS to a median 32x depth per donor and RNA-seq at a median 82.6 million reads per sample; sequencer models, compute hardware, and runtimes are not stated.
- Reproducibility gaps
- Protected GTEx data require dbGaP authorization (phs000424.v8) and raw sequence access runs through AnVIL; only non-protected V8 data are openly available on the GTEx Portal, and residual biospecimens are available on request.
Limitations (6)
- Trans-QTL discovery and characterization of sex- and population-biased cis-eQTLs remain power-limited: only 369 sex-biased (44 tissues) and 178 population-biased cis-eQTLs (141 eGenes) passed FDR ≤ 25%, with few large effects. · stated by the authors
- xCell enrichment estimates need cautious interpretation because validation data are scarce and reference profiles are limited in quality and number; external replication existed only for neutrophil ieQTLs, so other cell types relied on allelic expression checks. · stated by the authors
- Cell type interacting eQTLs and sQTLs (ieQTLs, isQTLs) are not true cell type-specific QTLs, since enrichment of one cell type may be correlated or anti-correlated with others; fully deconvolving cell type-driven cis effects is left as a future challenge. · stated by the authors
- Lower allelic heterogeneity observed for cis-sQTLs than cis-eQTLs may stem partly from weaker statistical power in splicing QTL mapping rather than biology alone. · stated by the authors
- Genotyped donors were 85.3% European American and GWAS analyses used European-subset QTL statistics, so how well enrichment and colocalization results transfer to other ancestries is not assessed. · noted in this summary
- Proposed mechanisms, such as EGR1 driving the CBX8 eQTL, CTCF disruption behind trans-eQTLs, and shared cell types explaining tissue sharing, rest on correlation, enrichment, or statistical mediation without perturbation experiments. · noted in this summary
Who should read:Read this if you use GTEx eQTL or sQTL data to interpret GWAS loci, prioritize causal regulatory variants, or study how tissue and cell type context shape genetic effects on expression and splicing.
Reproducibility ledger
Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.
Sign in to file a reproducibility report.
Related papers
- Local genetic effects on gene expression across 44 human tissues
François Aguet, Andrew A. Brown, … Stephen B. Montgomery·bioRxiv·Sep 9, 2016
- Distant regulatory effects of genetic variation in multiple human tissues
Brian Jo, Yuan He, … Alexis Battle·bioRxiv·Sep 9, 2016
- Cellular deconvolution of GTEx tissues powers eQTL studies to discover thousands of novel disease and cell-type associated regulatory variants
Margaret K. R. Donovan, Agnieszka D’Antonio‐Chronowska, … Kelly A. Frazer·Nature CommunicationsPeer-reviewed·Feb 19, 2020
- Exploiting the GTEx resources to decipher the mechanisms at GWAS loci
Alvaro Barbeira, Rodrigo Bonazzola, … Hae Kyung Im·Genome biologyPeer-reviewed·Jan 26, 2021
- A comprehensive catalogue of regulatory variants in the cattle transcriptome
Shuli Liu, Yahui Gao, … Lingzhao Fang·Nature GeneticsPeer-reviewed·Aug 11, 2022
- A systematic survey of human tissue-specific gene expression and splicing reveals new opportunities for therapeutic target identification and evaluation
Robert Yang, Jie Quan, … Hualin Simon Xi·bioRxiv·Apr 30, 2018