Credible Cockers has now closed and all posts are going on the Proper Cockers FB page. I've had permission to re post this explanation of the different types of DNA testing.
So much hearsay is being spouted on the back of the Wisdom Panel DNA test, the company themselves have stated it can't be used to prove pure breed.
I posted this in another group before and I thought it would be relevant here, because in the discussions I see not everyone know the difference between DNA tests for a disease, parentage or population:
(For the record, for me personally it doesn't matter one bit whether or not specific dogs are classed as a cocker or a mix. I just like people to judge the information based on facts. I do have some questions about the test, but I'll ask them in a seperate thread, because it can get confusing otherwise).
I saw the whole discussion about the dogs that are suspected sprockers. I do not want to start a discussion about that so please, please don't respond like this is about that.
I happen to be a biologist and I saw the questions and statements about the DNA testing, both in this group and other places. Some people seem to think there is a simple test to see the difference between two breeds, like there is for some genetic defects. The difference between those tests is what this thread is about.
With a DNA test for a genetic defect we are looking at one particular gene. We know (through prior research) what the correct gene's code looks like, we know a little piece of the gene or a region close to the gene that's different between the defective and the correct version of the gene, and when we look for that in a DNA test we can tell whether a dog is a carrier of that defective gene or not. The prior research is a lot of work and very technical, but this is all you need to know about a DNA test for a gene.
DNA test for parentage is a bit different, but very reliable. We look at some microsattelite markers, pieces of DNA that vary in length. It's simple, because the alleles an individual has must be explained by the alleles the suspected parents have. If they don't match on one marker it could still be the parents in some specific cases, but if multiple markers are off the chance is very very tiny. This kind of test is very reliable.
With a DNA test for breeds, we're not looking at a single gene. Not even at a couple. We're looking at the DNA in between genes. The non-coding DNA. We don't know if it does anything and if so what. The non-coding DNA has waaaay more variation than the coding DNA (that's because coding DNA gets fixed or an organism doesn't survive in the majority of copying mistakes, which doesn't hold true for the non-coding DNA). When we're looking at the difference between breeds we're looking at A LOT of variation. We're talking hundreds of thousands of little pieces of non-coding DNA that are used as a sample of the entire genome in big studies. Each of those pieces has multiple possible variants.
In a breed test, we say we know which variants are common in which breeds. That doesn't mean other variants don't happen in that same breed, it also doesn't mean those variants don't happen in other breeds. It is fascinating research, but it's not black and white. There's a huge overlap in variation between populations. The closer related populations are, the bigger the overlap will be. You might, for instance, see variant A most often in terriers and variant B most often in herding breeds, but that doesn't mean terriers can't have B and herding breeds can't have A.
However, when breeds are further apart, more and more differences can be found, and all those pieces of information together can tell a lot about how distant breeds genetically are from each other on average. But the thing with averages is, they are just that, an average. For instance, men on average are taller than women, but that doesn't mean there aren't plenty of women who are taller than some men.
Anyhow, when breeds are closer together it gets more and more difficult to genetically show a difference between the breeds, because there can be a big overlap in genetic diversity between those breeds. Commercial tests that tell you what breeds are behind your mixed breed dog, for example, are very unreliable because of this. Come time and more research they will improve, I'm sure, but they are fun, not hard science.
In other words, a test to see what breed a dog belongs to, is a completely different kind of test than one to see whether a dog has a defective gene.
I am not saying this test in very unreliable. I don't know this test. I have my doubts though and would not take it for truth without further research (I would suggest parent-offspring testing, that's very reliable). Spaniels used to be one big population, so much of their genetic variation will overlap between different spaniel breeds. Springers may on average show a certain set of genetic variants more often, but that doesn't mean it is impossible for cockers to share that set. Just like on average cockers are smaller, but some cockers are taller than some springers.
Also, we just assume the dogs originally tested were 'pure' cocker and 'pure' springer, whatever that may mean, so we don't even know for sure what the baseline is. We also don't know how working cockers differ from show cockers and how working springers differ from show springers.
It is of course very possible that things happened that shouldn't have (like is the case in many breeds), but this is not the place to discuss that. The only thing I wanted to do was educate you somewhat on how a DNA test for a genetic defect is different from a DNA test for a breed of origin, and because of their differences, their statistical reliability is very different. So please keep in mind these results have a chance (small, big, I don't know) to be incorrect. And please also keep in mind that this does not mean DNA tests for genetic defects are not reliable. It's apples and oranges.
