Nathan Fielder's Theranos Documentary Dropped Its Trailer. A $65 Dog DNA Test Once Ended With a Pug Being Euthanized.

A24 released the trailer this week for You Can See Everything, Nathan Fielder's documentary about Elizabeth Holmes, secretly filmed over three years and headed to theaters in October after premiering at Telluride.[1] Theranos is back in everyone's feed, and the lesson people keep drawing from it is a narrow one: that a lab result can be wrong.

The broader lesson is about what happens next — what a person does with a number they can't check.

There's a version of that in the pet aisle, and it's growing fast. The global pet DNA testing market was worth about $404 million in 2024 and is projected to reach roughly $719 million by 2030. Dogs make up a little over half the revenue. Breed profiling is the biggest slice; health and wellness testing is the fastest-growing one.[2]

Petunia

In 2018, three researchers — a veterinarian and bioethicist, a lab-animal specialist, and a canine geneticist — published a commentary in Nature under the title "Pet genomics medicine runs wild."[3]

The case at the center of it was a 13-year-old pug named Petunia. She had started having trouble walking and had lost control of her bladder and bowels. Her owners bought a $65 direct-to-consumer genetic test, which reported that she carried a variant linked to a degenerative neurological disease often compared to ALS.

They euthanized her.

The difficulty is what the result actually meant. As Dr. Lisa Moses, one of the authors, put it in follow-up reporting: research suggests only a small percentage of dogs who test positive for that mutation go on to develop the disease.[4] Petunia's symptoms had other possible explanations, several of them treatable in a 13-year-old dog.

The step that often gets skipped

Geneticist Elinor Karlsson, another of the authors, described the structural gap plainly. A study finds a variant that shows up more often in sick dogs than healthy ones. That's a starting point. The second step — testing whether that variant actually predicts illness in a new population of dogs — frequently never happens.[4]

The variant nonetheless makes it onto a consumer panel, where it arrives in an owner's inbox looking like an answer.

Human genetic testing has an oversight apparatus built around exactly this gap: validation requirements, clinical interpretation, and a genetic counselor sitting between the result and the decision. The pet version has almost none of it. The authors' argument was that it is too easy for companies to sell false hope, and that the market needs standards it currently lacks.[3]

What these tests are actually good at

This isn't an argument for skipping them. It's an argument for knowing which part of the report is load-bearing.

Reasonably solid: breed ancestry, which is what most people buy the test for anyway. Physical traits like coat and size. A handful of well-characterized single-gene conditions with a long research history in specific breeds — the MDR1 drug-sensitivity variant being the most useful example, because it changes real prescribing decisions.

Much softer: complex disease risk, where a single variant is one input among many, and predictions about behavior.

On that last one, the largest study of its kind is worth sitting with. Researchers sequenced 2,155 dog genomes and collected 18,385 owner surveys across 78 breeds, and found that breed explains only about 9% of the variation in behavior between individual dogs. For "agonistic threshold" — how easily a dog is provoked — breed was almost useless as a predictor. Even Labrador Retrievers, the breed least inclined to howl, had 8% of owners report howling.[5]

If breed itself only accounts for 9%, a behavior claim derived from a breed percentage is doing very little work.

What to do with a result that scares you

• Treat it as a screening result, not a diagnosis. A variant is a probability statement about a population, not a fact about your dog.
• Bring it to your vet before you bring it anywhere else — including into a decision about the dog's future. Ask what else could explain the symptoms in front of you.
• Ask about confirmatory testing through a veterinary diagnostic laboratory, and whether a veterinary geneticist should look at it.
• Ask what would change. For most reported variants, the honest answer is monitoring rather than treatment. That's still worth knowing, and it's a different thing from a prognosis.
• Don't make an irreversible decision on a cheek swab. Petunia's owners were trying to spare her suffering. They were acting on the best information they thought they had.

For $60 to $200, a dog DNA test buys a genuinely interesting ancestry breakdown and a handful of clinically useful flags. What it doesn't buy is a verdict — and the distance between those two is where the money is worth spending carefully.

---

Sources

1. Variety, "Nathan Fielder Drops Trailer for Secret Elizabeth Holmes Doc 'You Can See Everything'"
2. Grand View Research, "Pet DNA Testing Market Size And Share Report, 2025-2030"
3. Moses, Niemi & Karlsson, "Pet genomics medicine runs wild", Nature, 2018
4. NPR, "Questions Rise About Shortcomings Of DNA Tests For Dogs"
5. Morrill et al., "Ancestry-inclusive dog genomics challenges popular breed stereotypes", Science, 2022; summarized by the Broad Institute