Massive RNA Dataset Uncovers Hidden Genetic Switches Behind Gene Expression

For more than a decade, genome-wide association studies, or GWAS, have transformed the way scientists hunt for the genetic roots of disease. These studies have pinpointed hundreds of genomic locations, known as loci, associated with conditions ranging from breast cancer to heart disease to type 2 diabetes. Yet a frustrating gap has persisted between the discovery of these associations and a genuine understanding of what they mean. Most GWAS signals do not fall within the protein-coding sequences of genes themselves. Instead, they cluster in the regulatory regions of the genome, the stretches of DNA that act like dimmer switches, controlling how strongly a gene is expressed and how its RNA transcripts are spliced into their final functional forms. Without knowing which gene each regulatory variant actually influences, researchers are left with correlations they cannot explain and drug targets they cannot confidently pursue.

A new study published in Science, led by Peter Orchard, Ph.D., and Steve Parker, Ph.D., of the Gilbert S. Omenn Department of Computational Medicine and Bioinformatics at the University of Michigan, together with Laura Scott, Ph.D. of the University of Michigan, Laura Raffield, Ph.D. of UNC Chapel Hill, and Francois Aguet, Ph.D. of the Broad Institute, argues that the barrier to closing this gap has never been biological. It has been a matter of scale. The team’s analysis of tens of thousands of RNA sequencing samples from the Trans-Omics for Precision Medicine, or TOPMed, program reveals that genes are governed not by a single regulatory switch but by multiple independent ones, and that the second, third, and fourth switches only come into view when the underlying dataset grows large enough.

To appreciate why this matters, it helps to understand the machinery the researchers were mapping. Genetic variants that correlate with gene expression levels are called expression quantitative trait loci, or eQTL, while variants that correlate with how RNA transcripts are spliced are called splicing quantitative trait loci, or sQTL. These are the molecular bridges between the raw statistical associations produced by GWAS and the biological mechanisms that actually drive disease. Because an individual can carry one of three possible genotypes at any given variant, depending on which alleles are inherited from each parent, each genotype can be linked to a corresponding dosage of gene expression: low, intermediate, or high. Plotting these relationships across thousands of people allows researchers to measure how strongly a variant tunes the output of a gene, and that measurement is what converts a statistical signal into a testable hypothesis about mechanism.

Previous efforts to build these maps relied heavily on the Genotype-Tissue Expression project, known as GTEx, a landmark resource that catalogued gene expression across dozens of tissues. But GTEx and similar efforts ran into a ceiling. After a certain number of samples, researchers stopped finding new eQTL and sQTL signals for each gene, and the field largely concluded that most genes were regulated by one or perhaps a handful of independent variants. The new TOPMed analysis, which examined 14,324 RNA sequencing samples, including 6,454 whole blood samples and 1,291 lung samples, shows that this ceiling was an artifact of sample size rather than a true feature of the genome. After performing eQTL and sQTL analysis across the TOPMed cohorts, the team mapped tens of thousands of signals with high resolution, and the number of independent signals per gene kept climbing.

“The more samples you add, the more e/sQTL signals that you detect,” Parker explained. “I think NHLBI has invested very smartly in TOPMed. Although you saturate discovery of main effects after a few thousand samples, the number of independent e/sQTL signals just keeps on growing and growing, which is really cool to see.” The implication is striking: even in a tissue as heavily studied as whole blood, most genes harbor additional regulatory switches that smaller studies simply cannot resolve. These secondary, tertiary, and quaternary signals are not redundant copies of the primary one. They represent genuinely independent layers of control, each capable of modulating gene expression or splicing in its own way, and each potentially relevant to different diseases or environmental contexts.

The genetic diversity of the TOPMed cohorts proved to be just as consequential as their size. Roughly one in six of the eQTL signals the team detected in whole blood came from variants that are rare or absent in people of European ancestry. Yet almost none of these signals, just 1.3 percent, appear in the four large eQTL resources built predominantly from European-ancestry samples. This means that the reference databases the field has relied upon to interpret GWAS results are missing a substantial fraction of the regulatory architecture present in human populations, particularly in groups that have historically been underrepresented in genomic research. “Another powerful aspect of larger-scale studies is the ability to identify QTL for rarer variants, which tend to have larger effects,” said Aguet. “As further such studies are conducted across more tissues and cell types, we will likely find a lot more of these effects and increase our understanding of their contribution to complex traits.”

To demonstrate the practical payoff of this richer map, the researchers fine-mapped GWAS signals from the UK Biobank, one of the largest and most extensively studied biobanks in the world, against their new eQTL catalog. The pattern was unambiguous: the more eQTL signals available for a gene, the more GWAS associations could be confidently connected to it. Of the 10,611 GWAS signals the team was able to link to a specific gene, roughly two-thirds involved one of the secondary switches rather than the primary regulatory signal that smaller studies would have detected. In other words, the majority of disease associations that previous efforts struggled to explain may have been pointing at regulatory mechanisms that were invisible to the datasets of the time.

“With GWAS, you find regions of the genome that correlate with the probability of having a particular disease, like type 2 diabetes for instance,” said Orchard, the first author of the paper. “You know there’s a correlation, but you just don’t know why. With the eQTL, you’re trying to bridge that gap and add more context to the associations between variants and the traits we care about.” That bridge has enormous practical value. Parker noted that genetic evidence supporting a drug target is one of the strongest predictors of clinical success for therapeutics, increasing the probability of success more than two-fold. When a GWAS signal can be traced through an eQTL to the specific gene whose expression it modulates, that chain of evidence elevates the gene from statistical suspect to credible therapeutic target. Scott added another layer of biological nuance: “Genetic regulation of the expression levels of one part of a biological unit’s machinery, such as a regulator of a cellular stress response, can affect the expression levels of other parts of the same biological machinery.” Gene regulation, in other words, is not a collection of isolated dials but an interconnected network, and mapping more of its nodes reveals how perturbations propagate through cellular systems.

The findings also carry a clear message about where the field should go next. The authors emphasize the importance of collecting tissue samples beyond whole blood, since regulatory logic differs across tissues and many disease-relevant mechanisms operate in organs that blood samples can only partially illuminate. Raffield noted that even within whole blood, TOPMed still holds many undiscovered signals at most genes, which will almost certainly provide further links between GWAS loci and their likely mechanisms. She also cautioned that many non-European ancestry populations remain underrepresented in the current dataset despite deliberate efforts to be more representative than earlier projects, meaning that important ancestry-differentiated signals may still be missed, in blood and certainly in other tissues. “These issues will continue to be addressed in future work from TOPMed,” she said.

“Science moves forward with bigger and better data, always,” Parker reflected. “It’s the equivalent of getting a fancy new microscope, except we’re not seeing things in higher resolution in a microscopy image, we’re seeing things in much higher resolution from a genetic association with gene expression perspective.” The study, titled Cross-cohort analysis of expression and splicing quantitative trait loci in TOPMed, was supported by the Foundation for the National Institutes of Health, the National Heart, Lung, and Blood Institute, the National Cancer Institute, the National Institute of Environmental Health Sciences, the COPD Foundation, the American Asthma Foundation, and the Evans Medical Foundation. Its central lesson is deceptively simple: the genome’s regulatory switches were never missing, only unseen, and the instrument needed to reveal them is not a new technology but the patient accumulation of data from populations diverse enough to reflect the full breadth of human genetic variation.

Subject of Research: Mapping expression and splicing quantitative trait loci across diverse human cohorts to link genetic variants to gene regulation

Article Title: Bigger data reveals the second, third and fourth switches controlling a gene

Article References: Bigger data reveals the second, third and fourth switches controlling a gene. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: GWAS, eQTL, sQTL, gene expression, TOPMed, GTEx, genetic regulation, RNA sequencing, precision medicine, UK Biobank, genetic diversity, drug targets