A comprehensive new genetics study of more than a thousand golden retrievers has unveiled a cluster of behavior-linked genes that appear to shape not only how dogs behave, but how humans do, too.
In a paper published in PNAS, a team of researchers from the University of Cambridge (and collaborating institutions) used genome-wide association studies (GWAS) to probe the roots of canine temperament, analyzing more than a dozen behavioral traits across nearly 1,200 adult golden retrievers. Their goals? Clarify the genetics of common dog behaviors while trying to find out if the same genes are relevant to us?
“The findings are really striking. They provide strong evidence that humans and golden retrievers have shared genetic roots for their behavior. The genes we identified frequently influence emotional states and behavior in both species,” study lead and University of Cambridge researcher Eleanor Raffan, PhD, said.
More specifically, of the 18 strongest genetic candidates flagged in dogs, two-thirds also show strong ties to psychiatric, temperamental, and cognitive human traits. And that overlap, Raffan and her colleagues argue, might reveal more about the emotional engines that drive dog behavior than any previous research has.
Why Golden Retrievers?
Purebred dogs offer a huge advantage when it comes to research. Generations of selective breeding have produced animal populations with high genetic similarity and long blocks of linked genetic material. And that structure makes it easier to spot behavior-related variants that, in humans, would require much larger sample sizes.
This new research draws on the “Golden Retriever Lifetime Study,” a massive ongoing cohort that tracks genetics, health, lifestyle, and owner-reported temperament.
To measure behavior, researchers used the C-BARQ questionnaire, a tool that distills everyday behaviors into 14 quantifiable traits.
The dogs in this research project were all between ages 3 and 7, an intentional choice to weed out the “noise” of puppyhood along with the cognitive decline that comes with old age.
What They Found
The team identified 21 genetic loci linked to behavior: 12 reaching genome-wide significance and nine “suggestive” hits that they say should warrant further study. These links covered eight major traits, including dog-directed fear, stranger fear, non-social fear, aggression toward other dogs, energy levels, separation-related problems, trainability, and touch sensitivity.
Some loci contained just a single nearby gene. While others stretched across 4.5 megabases and held more than a dozen candidates. The researchers then examined human orthologs of the nearest genes for each canine finding, combing through 190 psychiatric, cognitive, and personality traits.
Twelve genes lit up. And most of them did so repeatedly.
Shared Genes, Shared Emotional Rules
The cross-species parallels revealed some startling emotional themes. For genes tied to fear and aggression in golden retrievers, the top human links were often neuroticism, anxiety, irritability, mood swings, depressive symptoms, and sensitivity to social stressors.
One of the clearest examples centered on PTPN1, the gene closest to the major dog-directed aggression locus. In human studies, PTPN1 revealed ties to intelligence, cognitive performance, educational attainment, and major depressive disorder.
Another gene, ROMO1, revealed a connection to trainability in dogs but to intelligence, cognitive performance, and depression in humans.
Even seemingly mismatched traits showed a stronger consistency. ADGRL2, tied to trainability in golden retrievers, surfaced in humans in connection with guilty feelings, irritability, and emotional sensitivity.
In other words, what looks like “stubbornness” or “high drive” in a dog could just be the behavioral expression of temperament-shaping neurobiology – a biology we share.
What It Means for Dogs
The findings challenge some long-held beliefs about canine behavior – especially the idea that fear, aggression, or reactivity are learned behaviors.
Instead, the findings tend to suggest that some dogs might be genetically predisposed to intense emotional states that prime certain behaviors long before a trigger appears.
For traits like dog-directed fear, dog-directed aggression, and non-social fear, the human parallels—neuroticism, anxiety, depression—suggest these dogs might live in a baseline of heightened emotional arousal. That certainly makes behavior modification more complex. But it opens the door to potentially more humane approaches. Clinicians might just need to treat these dogs as emotionally vulnerable, not merely “misbehaving.”
While the study wasn’t designed to draw conclusions about human psychiatric genetics, the cross-species analysis does offer an unusual piece of “orthogonal evidence.”
But maybe more importantly, the effect sizes that the researchers noted in the dogs appeared to be larger than in human studies, simply because purebred canine genomes make big effects easier to detect. That gives researchers clearer targets and could elevate some of these genes as priorities for mechanistic study.
A New Model for Behavior Genetics?
Researchers already rely on dogs as models for compulsive disorders, autism-related traits, and age-related cognitive decline. This latest research proposes a new category: emotional and temperamental architecture itself.
The promise isn’t that dogs perfectly mimic human psychiatric conditions. They can’t. But they might just illuminate the biological scaffolding that supports traits such as fearfulness, mood vulnerability, and emotional reactivity.
And for dog owners, the work offers a gentler reframing of difficult behavior. Sometimes, a dog isn’t being “bad.” Sometimes, they’re wired a little differently. Just like us.
Further Reading
Dogs Can Smell Our Stress And It Burns Them Out
