Here’s what you’ll learn when you read this story:

• Eye color comes down to a pigment called melanin. The more melanin your irises produce, the darker your eyes.
• Your genes are the main thing controlling how much pigment you make. But your DNA blueprint doesn't always match what you see in the mirror.
• A new study found that extra genes can quietly dial melanin production up or down, overriding the primary eye-color gene and producing surprises.

Human eyes come in a kaleidoscope of colors—from the palest blues of the sky to the deepest browns of the earth. Some change color in shadow or sunlight, and others even glimmer with more than one shade.

In humans, eye color is a trait that is highly polygenic, meaning it’s influenced by more than one gene. Genes have interactions with other genes and their environments that control their expression, and how they’re expressed determines the appearance, or phenotype, of the eye. The way these genes are turned on or off determines how much melanin the iris produces. Higher levels of melanin in the iris create darker eye colors, and vice versa.

So which specific bit of DNA is really running the show when it comes to melanin levels in your eyes? Scientists point to a tiny genetic variation in nucleotides (which make up DNA and RNA). Basically, it’s just a single-letter swap in the genetic code called a single nucleotide polymorphism, or SNP (pronounced "snip"). The one that matters most for eye color goes by the deeply uninspiring catalog name rs12913832. Think of it as the eye-color switch that evolution never bothered to label properly.

That switch sits inside a gene called HERC2, whose main job is to boss around a neighboring gene called OCA2. And despite its unimpressive name, OCA2 is an especially important gene for eye color: it’s the one that actually helps produce melanin. The HERC2 switch comes in two versions, known as alleles (gene variants): A and G. If your genetic code carries the A version, HERC2 tells OCA2 to crank up melanin production, which will make your eyes turn out more brown. But if you carry the G version that signal gets dialed way down, so less pigment is made, and your eyes will trend toward blue.

But the way you look isn’t always determined by which genes you have, or as scientists would put it, phenotype isn’t always determined by genotype. While eye color phenotypes are usually lighter when there is a GG genotype, and darker with AA or AG genotypes, a team led by researchers from the University of Toronto—who used a genetic database to compare the eye colors of nearly 5,500 people to their genotypes—discovered that’s not what always occurs. Some subjects, for instance, described their eye color as being something other than blue despite the expression of their OCA2 gene being “switched off” enough to result in blue eyes. And some individuals with increased OCA2, which should result in brown or hazel eyes, had blue eyes. In other words, the effects of having the A or G alleles of the HERC2 switch aren’t as clear-cut as scientists assumed. The AA, AG, and GG genotypes alone can’t be the only determinants of eye color. The logical conclusion is that some other genetic variants must be having an effect on melanin and pigmentation.

“Despite the heavy reliance on this single SNP in forensic and anthropological models, exceptions in phenotype suggest additional modifiers. Our genome-wide approach successfully identified several SNPs that modify eye color in these ‘non-concordant’ individuals,” the researchers said in a study recently published in the journal Scientific Reports.

When the researchers went looking for what was overriding the main eye-color gene, they found a handful of supporting players—genes that are involved in other steps of the melanin-production process. In people with GG genotypes whose eyes were darker than expected, the culprits were genes that either help manufacture melanin directly (TYR and TYRP1) or move raw materials into the tiny cellular compartments called melanosomes where pigment is assembled (SLC24A4, SLC45A2, and TSPAN10). Variants in these genes effectively compensated for the reduced OCA2 signal, boosting pigment production enough to push eye color from blue toward brown.

The mirror-image pattern showed up in people whose DNA predicted brown eyes, but who ended up with lighter ones. Here, the disruption worked in reverse. Instead of extra genes boosting pigment, they were quietly undermining it. One key player was a gene called IRF4, which normally helps activate the pigment-production enzyme TYR. In some individuals, a variant in IRF4 dialed down its activity, which meant that melanin output dropped. The result was that instead of being brown, eye color shifted toward green or hazel. The researchers also found that OCA2 (the same gene the HERC2 switch is supposed to be controlling) had its own independent weak spots. Two variants within OCA2 itself reduced its function regardless of what HERC2 was telling it to do, effectively giving the main eye-color gene a second way to underperform. A third modifier, TYRP1, turned up here too, interestingly appearing in both the darker-than-expected and lighter-than-expected groups — suggesting it can push pigment levels in either direction depending on genetic context.

“The identification of shared and background-specific loci underscores the polygenic and regulatory complexity of iris pigmentation,” the researchers said. “Our findings provide new insights into how moderating genetic factors contribute to subtle (and sometimes more pronounced) deviations from expected eye color.”

It turns out that your eye color is a lot more complicated than it seemed.