How Accurate Are Those Pet DNA Tests?

How accurate are those pet DNA tests really? Boston's CBS WBZ Investigative Team decided to put dog breed tests to the test, and they got pretty surprising results. 

According to Elinor Karlsson, a geneticist with the Broad Institute and UMass Chan Medical School, and who runs a lab called Darwin's Ark, where she digs into pet DNA science states, "I personally do have concerns about the fact that, from a consumer standpoint, you don't always know what you're getting when you work with those companies. There's not a lot of rules in this space." 

So last year, the WBZ I-Team checked up on commercial DNA testing companies by sending a sample from a New Hampshire pet owner's human cheek to a company called DNA My Dog. The results showed she was part bulldog! 

When the company was contacted, the DNA My Dog service director responded that those results would not be possible on a human sample. So, the WBZ I-Team decided to check it out again. Reporter Christina Hager swabbed her cheek and sent her samples to three companies. 

One company, Orivet, reported that the sample "...failed to provide the data necessary to perform the breed ID analysis." Another Wisdom Panel sent a message saying the sample "... didn't provide...enough DNA to produce a reliable result." 

However, the previously used DNA My Dog company linked the reporter's DNA sample to dog breeds for the second time! The company reported that Christina Hager was 40% Alaskan Malamute, 35% Shar-Pei, and 25% Labrador. When the Boston WBZ I-Team contacted the Toronto-based company again, DNA My Dog did not respond despite repeated attempts. 

At the risk of sounding incredibly sarcastic, what a surprise! It is outlandish that these DNA companies cannot decipher human DNA from canine DNA. Yet, are we supposed to "trust" their results after paying them $80 for a DNA canine test?

All of the above makes me even more skeptical of this DNA market based on one or more results I know about. For example, a rescue dog that appeared to be a miniature Wolfhound had DNA returned as 50% Standard Poodle, 46% Dutch Shepherd, and 4% American bulldog. From the results, it had no Irish wolfhound DNA. 

Why am I distrustful of these particular results? It may have had no Wolfhound in it, but I find it hard to believe the results because of the 46% Dutch Shepherd DNA, which is highly unlikely. The Dutch Shepherd is a rare breed in the Foundation Stock Service (FSS) stages of recognition within the AKC and is in the Miscellaneous Group. It is so rare that there are approximately 11,000 in the world. Of course, it is possible that this commercial "Doodle" puppy mill obtained one Dutch Shepherd specimen and combined it into its substandard crossbreeding operation, but I find that improbable. How do I know it was a commercial puppy ill operation? Because the dog had paint on its back, the color of which identified who it was bred to. 

The point is that you would require more than one Dutch Shepherd breed specimen to consistently reproduce subsequent generations. Yes, it is far-fetched but still possible to obtain numerous Dutch Shepherds, mainly if they were imported from Europe, which segues into an even more critical discussion about European Breeders indiscriminately shipping puppies to the United States, willy-nilly. I deviate here but will tell my European counterparts that dogs are not livestock. Even more, we no longer are in the golden ages of yesteryear when conscientious, ethical, private American breeders throughout the U.S. were interested only in whatever breed's welfare, adhering to the custodial principle. Sigh, an article that will have to wait for another day. 

In closing, I am dubious to the point of simply not believing in these Canine DNA testing company results. One pet owner received results that included as many as 14 breeds, including Chihuahua, in her medium-to-large-breed rescue. Dr. Lisa Moses, a veterinarian and bioethicist at Harvard Medical School, believes the science needs to be more precise because there are no official definitions for breeds. She says there are no exact genetic codes to match them. "There isn't necessarily a gold standard answer for what your dog is... A breed is something that we've decided, which is based upon essentially the way a dog looks," said Dr. Moses. "But that doesn't necessarily mean that we're going to know what their genes look like."

I find it all to be nonsense. I believe in the DNA Health research into genetics to identify canine cancer and other health risks. The above bogus science can only do harm in the public's mind. 

Read the entire short DNA article here on CBSNews.com.