In 2010, the US National Institutes of Health launched an ambitious project: collect tissue samples from hundreds of deceased donors, sequence the RNA from each tissue, and build a comprehensive map of how genes are turned on and off across the human body. More than a decade later, the Genotype-Tissue Expression (GTEx) program has transformed understanding of tissue-specific gene regulation, and the data is still yielding discoveries.
The final GTEx dataset (V8) contains genetic data from 838 postmortem donors and RNA sequences from 17,382 samples across 54 tissue sites and two cell lines. Each sample captures the transcriptome, the full set of RNA molecules produced by the genome, in a specific tissue from a specific individual. The result is a resource that allows researchers to ask a question that was previously unanswerable at scale: how does the same genome produce a heart, a liver, and a brain?
What GTEx revealed
The most immediate finding was that most genes are not uniformly expressed. The majority show tissue-specific patterns. Some genes are active only in the liver, others only in the brain, still others only in the testes. But the scale of GTEx allowed researchers to go beyond binary on/off classifications to measure the quantitative variation, how much more a gene is expressed in one tissue compared with another, and how that varies across individuals.
This quantitative variation turned out to be the key to understanding expression quantitative trait loci (eQTLs), genetic variants that influence how strongly a gene is expressed. Before GTEx, eQTL studies were largely limited to blood, because blood is easy to collect. GTEx showed that many eQTLs are tissue-specific. A variant that boosts expression of a gene in the liver may have no effect in the brain, a finding with direct implications for drug development, where the target tissue matters.
GTEx enabled the development of PrediXcan, a statistical method that uses an individual’s DNA sequence to predict their gene expression levels in specific tissues. This tool has been used to identify genes associated with bipolar disorder, coronary artery disease, Crohn’s disease, rheumatoid arthritis, and type 1 diabetes, conditions where the relevant tissue (brain, artery, gut, joint) is not accessible in living patients.
The scale of the resource
The GTEx program was originally funded by the NIH Common Fund with a 5-to-10-year horizon. The project formally transitioned from Common Fund support in 2019, but the data continues to be used and cited. The tissue bank at the Broad Institute still holds unused biospecimens available for new research.
The dataset has been integrated into the UCSC Genome Browser, allowing anyone with an internet connection to visualize gene expression patterns across tissues. The GTEx Portal provides searchable access to expression levels, eQTLs, and splicing patterns.
Lessons for the next decade
The GTEx model, large-scale, publicly funded, open-access biospecimen collection, has influenced subsequent projects including the Human Cell Atlas and the NIH Bridge to Artificial Intelligence program. The challenge now is moving from bulk tissue RNA sequencing to single-cell resolution, where each cell’s transcriptome is measured individually, revealing the cellular heterogeneity within each tissue that bulk sequencing averages out.
Projects combining GTEx-style tissue mapping with single-cell methods are already underway. The goal is a reference map of human gene expression at cellular resolution across all major tissues, a resource that would make the GTEx atlas look like an early sketch of a much more detailed picture still to come.