Friday 6 September 2013

Every breed... every test... every lab.. Today's useful link

Click to enlarge
Want to check what DNA tests are available for your breed (or crossbreed) and which labs worldwide offer them?

You can now, here.

Thank you to PennGen at the University of Pennsylvania, the WSAVA and Waltham. I couldn't catch it out on a test-ride in terms of breed-specific tests offered - although I found the drop-down menus awkward to use and the info takes time to load on luddite internet connections.

But this is only a small niggle. It's a great resource.

30 comments:

  1. Very useful. Not totally comprehensive as many of these tests are also offered through referrence labs such as Idexx and Axiom at cheaper rates but pretty good never-the-less. I tried several of the disease info links and there was no info available but that will come I guess. If they want to make this really superb then rating the labs would be a good add on (accreditations, customer feedback) as would a traffic light system of disease/test relevance as some are extrememly rare conditions in the breeds mentioned but which have been used as models for human disease. (eg in my breed, ESS familial Nephropathy and Shaking puppy syndrome just aren't seen in 'real life' at present).
    VP

    ReplyDelete
  2. My breed isn't listed.

    ReplyDelete
  3. I would much prefer to see development in a whole genome direction, as in 23andme.com (for human genetic testing). I hate to see breeders confronted with a smorgasbord of tests all based on the same material (DNA), with each lab asking for its own fee...some of the tests (eg, narcolepsy) pretty irrelevant because the condition is rare, and some of the laboratories involved churning out results that are, in fact, incorrect,

    ReplyDelete
    Replies
    1. Jennifer are you a laboratory professional? Why do you refer to labs churning out results that are incorrect? If a laboratory is operating in a professional capacity, they will be obliged to have systems in place to ensure that the tests and results are correct. Otherwise, why are they in business? Internal quality control systems plus external quality assurance schemes help labs keep their standards in place. Accreditation schemes audit and monitor their processes too. Perhaps you have some examples of incorrect results you could share?

      Delete
    2. I am a scientist who has done some dog breeding. I've had trouble with test results showing a dog carried the (recessive) gene for chocolate color. He was bred to a chocolate bitch and there were no chocolate or chocolate point pups in a litter of 10. My math is not bad . . . I calculated the odds at around 1:1000 of this happening. We sent a DNA sample to another lab and the results came back saying the dog did not carry the chocolate gene. I have discussed this over dog forums (forae?) with other breeders, and I am not alone in getting results that were demonstrably wrong. The quality schemes in some labs are not what they should be. When you have a large number of laboratories doing a large number of tests, it is almost inevitable that there will be quality control problems.

      you're right, I shouldn't have used the word 'churning'

      Even an error rate of a few percent on a serious recessive condition is a BIG worry. It would be horrible for breeders who are trying to do the right thing to find out that the dog they though was clear was, indeed, a carrier. Should this happen with a late-onset disease, chances are that it would not be discovered until after old age removed the dog from the breeding pool...or it might go undiscovered for two or more generations if the dog wasn't used much and the practice of 'clear by parentage' was used to determine carrier status.

      Delete
    3. p.s. color tests are a good canary for the mineshaft. It's no big deal if the color inheritance comes out other than predicted by tests. It's a big deal if some serious health problem comes out with a false negative. False positive is also a concern, but much less damaging to the dog population.

      main point . . . before we subscribe to genetic testing we need to ensure that excellent quality control is in place. I'd love to see indications that this is true. Until such time as I see evidence, I remain, as always, skeptical.

      Delete
    4. You are right, Jennifer. But unfortunately, labs with dubious records are the ones that often very freely threaten litigation if you name them.

      One of those labs is listed on the PennGen site, I see.

      Unfortunately, it's a minefield. What can breeders do? Look for peer-reviewed papers supporting the test. And another good source are the breed fora. If there's a lab producing dodgy results, there will be discussion about it there.

      Jemima

      Delete
    5. Thanks for clarifying Jennifer.

      Some suggested questions for owners to ask potential screening laboratories are:
      1) What standards of practice do you follow?
      2) Are you registered with an appropriate accreditation body?
      3) Can you provide evidence of your QC records, including the corrective and preventative action if a result falls out of testing range?
      4) Do you participate in any external quality assurance schemes and what steps do you take if any results fall outside of the ranges?
      5) Are the staff undertaking the tests appropriately qualified scientists registered with any professional bodies?

      The more 'yes' answers, the better! I would not touch a lab with a barge pole if they did not display there QC and QA documents. Labs will always get things wrong - a defect rate will be proportional to the sensitivity and specificity of the test plus the sheer volume of testing, however, if they are a professional organisation there will be protocols in place to address this and ensure that people understand the sensitivity and specificity of the result they have. They should have records of defect rates/errors etc. within their quality management system. And yes, they SHOULD have a QM system as minimum. If not, run a mile...

      The dog world lacks professionalism at times and my fears for genetic screening are that people will become too attached to the outcome of a result and place too much emphasis on it - they will take it too literally. Results have to be assessed in context with a whole host of other stuff too...hope that makes sense, I'm rushing to get out to take t'dogs for a walk between rain showers.

      Delete
    6. Thank you Best-Practice-Lab-Anon....

      ;-)

      Jemima

      Delete
    7. Best-Practices-Lab-Anon
      Those are good suggestions for choosing which lab to use for a test, presuming there's a choice.
      It all becomes very messy if you're looking to use a stud dog and he has been tested. How are we to know if XYZ Lab had a high (or very low) error rate five years ago when the dog was tested?
      We really need either government or some interest group (perhaps a veterinarians trade group?) to provide external monitoring and certification.

      Delete
    8. Because the lab should keep accurate records Jennifer if they are professional and working to Good Laboratory Practice (GLP). Otherwise, don't use the Lab. And you have to start somewhere....at present, it isn't set up in the best interests of the users.

      I agree with regard to external regulation. Accreditation should ensure these basic GLP principles, which will all be written down in the standards that the laboartory is benchmarking to. Accreditation requires internal and external audit processes, the latter requires regular inspections from the Accreditation body. If they fall short, they fail the Accreditation process. They then lose business because only accredited laboratories will get work. Well, that's the theory...

      I work in research - if we lose accreditation status, we lose work because part of the company's evaluation process to vet a lab's capabilities is that we can provide evidence that we work to a set of standards which are measurable, appropriate and that we are committed to continually improving our processes based on empiricism. Therefore, we will lose our jobs if there is no work to fund our salaries. It's a pretty powerful motivator.

      There should be a labortaory reference list available (where, not sure ? KC ? vet group) of ONLY the accredited laboratories that provide DNA testing. If a lab is not on that list, don't use it. If a result is obtained from a lab that isn't on the list, the result is invalid.

      The breed clubs should ensure that all breeders only use suitably accredited bodies for any DNA/genetic screening. If we get serious about quality issues and start asking the right questions, hopefully people will catch on.... because they WILL understand. There has to be pressure from users.

      Delete
  4. Please change the niggle part. Although there is no human racism at all, I'm pretty sure most people will instantly think of the 'n' word. lol

    ReplyDelete
    Replies
    1. Oh for goodness' sake! Niggle is a perfectly acceptable word first recorded in English in 1599! I'm sure most people are far too sensible to confuse it with any other word that begins with the same three letters.

      Delete
    2. Oh SkyArk, you do make me laugh! It does seem to be lost on most people your sarcasm and left field humour! Looks like Mary fell for the trap though ;)

      Delete
    3. Hey I agree wholeheartedly. If I had my way no words would be censored, regardless of (or lack of) association, aside from malicious libel. Its just the reality and people get easily offended.

      It reminds me of the time I was searching for an answer to not getting hot in bed because it was making it impossible to be comfortable with a blanket. I even had to take my cloths off. As you can imagine, most people associated it with something else, and it would get off-topic fast.

      Its a pain, but its a reality of life that we gotta watch out for. If she keeps it, its fine by me, I'm just giving a little warning to what Americans might associate it with.

      Delete
    4. I don't believe Americans are that stupid (despite the case the other year when a man lost his job for using the word 'niggardly'. He was later vindicated, but didn't get his job back.)

      And why would people change the subject about taking clothes off in bed, especially when hot? A recent survey found that about a third of the UK sleeps naked.

      Delete
  5. Every breed? They don't even have Rottweilers on the list.

    ReplyDelete
  6. People have found a few niggles (not a bad word!), like lesser known breeds with similar names getting mixed up but when they have pointed out the errors to the developers they've been pleased to get feedback and have made ammendments. It's a work in progress and everyone can help :-)
    VP

    ReplyDelete
    Replies
    1. The site doesn't work at all for me. I scroll to specie and breed and it just sits there. It's like someone forgot to put in a SEARCH button to initiate the search. I've emailed them twice but they don't seem interested.

      Delete
  7. Its says "For mixed breeds select closest similar breed" so all these doodles and poos etc, still don't show what problems they have, and are developing themselves.........or are hiding under the Hybrid vigour lie. Sorry this is not a useful site.

    ReplyDelete
    Replies
    1. As if, somehow, you are a breeder of doodles...


      Delete
    2. Hybrid vigour lie!? Jemima you do well to let these comments get published! ;)

      Delete
    3. It amuses me that there is a huddle of Hate-Jemima-Anons that on principle believe that anything I say is good must be bad. And vice versa.

      Actually, the phrasing isn't great on the PennGen site. Selecting the "closest similar breed" could be less than useless. Much better if it stated that, for cross or mixed breeds, you should select the known breeds that went into the mix.

      Jemima

      Delete
    4. The first round of Labra-Doodles were actually purpose bred dogs. In keeping with the mantra of the Kennel club and registries these dogs were in fact designed by humans to initially serve a purpose as a guide dog with hypoallergenic properties that would be better suited for clients with doggie dander allergies.

      It was the naming of the dog that ruined it. And the Marketable name caused a rush of breeders to jump on board and make the same mix of dog.

      But, if they did go back to basics and try to reign in the standards and design a guide dog regimen to set a baseline of measure for guide dog abilities they would be back on course for creating the foundation for a breed with a more fitting name...

      The Austrailian Guide Dog.

      Just like our ancestors used mixing breeds to better meet a purpose... The creators of the Labra-Doodle did in fact do the same.

      Modern thinking about innovation exploitation for monetary gain ruined all the potential for them. But, its literally not too late to turn it around. But, they need to organize and get away from the naming convention of slamming the names of the foundation breeds used in the outcross together. If they can do that... then the "Australian Guide dog" can emerge from this a an impressive story if they actually repair all the mess they caused with un-researched matings.

      Delete
  8. In my opinion, the single most important test you can perform -- should work on all breeds and most species -- is measurement of homozygosity in the MHC (DLA for dogs). This provides a useful index of the extent to which inbreeding has compromised the genetics of the immune response system. With wild populations that have gone through a bottleneck, eg. island fox populations that have recovered after being decimated by an epidemic, it is used to show whether or not balancing selection has allowed recovery of genetic good health despite the narrowed gene pool. This database does not seem to include any such tests. If they are included, their identifying tag is not clear. Thumbs down for the tool.

    p.s. the Canadian poodle research referred to in the previous blog post have done a good job working the DLA (~MHC) tests. It may be that other rehions of DNA are just as, or nearly as, important as DLA. So a whole genome approach remains the preferred option.

    ReplyDelete
  9. For genome-wide testing (diseases, conformation, homozygosity of an individual - and the homozygosity of the breed as well when at least 30 dogs of that breed has been tested) check MyDogDNA test and database by Genoscoper: www.mydogdna.com.

    And no, I don't get any money for advertising this. ;-)
    I just find the whole idea wonderful and extremely useful. I have tested both of my Pharaoh Hounds and will test the rest of the pack as well.

    ReplyDelete
    Replies
    1. Post to come on this shortly...

      Hanne, would be interested in hearing more about your experience with MyDogDNA if you'd be happy to email me privately: jem@pedigreedogsexposed.com.

      Jemima

      Delete
  10. So the BBC dumped you and now have Ruth, Steve and Kate to show the world of dogs, so perhaps those overpaid execs and bloated editors have now learned a lesson care of Ofcom?

    ReplyDelete
  11. http://www.akcchf.org/canine-health/your-dogs-health/major-histocompatibility.html

    At the end of this article, it suggests that testing breeding stock for auto-antibodies to the thyroid, could prevent breeding from dogs that will later develop an auto-immune disease. Auto-antibodies to the thyroid mean that the immune system is already compromised.

    What I'm not sure of is, how to go about finding the test? Not that I want to breed, but it could be like an early warning system to an impending AI.

    ReplyDelete
    Replies
    1. I would have thought that just because you have a measurable titre of autoantibodies to the thyroid gland, it does not necessarily mean that you will develop an autoimmune disease. It would depend on the normal range of the population tested as to what is considered 'positive' or 'negative' as an indication of disease progression. So, in that study, which breed of dogs were tested and what were the longterm outcomes?

      Human beings walk around with detectable levels of autoimmune disease antibodies such as anti-microsomal antobodies (thyroid) ANA antibodies but aren't clinically in a state of disease, for example.

      Delete