Hello I thought people might be interested in hearing about my chat with the kennel club genetics chap and this new Estimated Breeding Value system that has now come online for labradors and a limited number of other breeds. First thing he did was to show me exactly how the software worked. It would be useful for those interested to log into mykc.org and look at it with a dog you know about perhaps a well known / well used stud dog and then a dog with far less data to show the main differences between them. The hip and elbow lines are separate and run from minus (red) to positive (green) numbers. The rating that they provide for each dog is calculated not just using the dogs own hip and elbow scores if available but progeny, siblings, parents, g parents and beyond. For a dog with few relevant dogs included in their calculation, the amount of certainty in the number is less and therefore the number has a greater chance of being incorrect so is shown with a higher spread on the bar. If you have a dog that has a lot of hip scored close kin, their number will be more certain and shown the "spread"shown will be far less. Its useful to look at a couple of famous dogs. He showed me a famous guide dog stud dog. you can see how his own hip score wasnt great but his progeny have worked out very well and thereforeit can be said that there doesnt appear to be a genetic reason for his poor score and it is more likely to stem from envirnmental reasons. so his own hip score is fairly poor but his ebv is far better. Then you have ftch craighorn bracken, a dog quite heavily used despite a 1 grade for his elbows. until recently, this wasnt a number that put many very good breeders off from using a stud dog. however if you look at craighorn bracken, youll see his ebv number for elbows is very very poor which shows that this 1 points to strongly genetic problem rather than an environmental factors Because both dogs were so well used and enough of the resulting pups have themselves been scored, the ebv provides a far more well rounded and easy to understand opinion on a dogs genetic soundness. As I see it, one of the problems with this scheme is that if youre looking to buy a pup with a good ebv that has a statistically high level of accuracy, young rarely used dogs arent for you. this is of course always going to be more difficult with the bitch who isnt going to have a high number of offspring who are tested. If you know what youre doing and spend a deal oc time analysing hip and elbow scores you have had the ability to consider all of these different aspects yourself without this ebv scheme, but for those less well adept at doing this, it's fantastic plus it removes quite a lot of work. In my opinion this is just the sort of thing that the kennel club sbould be spending its time doing. its not going to bring an end to these diseases on its own but it leaves less places for people who dont test to hide (as theyre dogs will have a score even though its unlikely to be as statistically accurate as having all the dogs scored), it also helps to isolate the genetical reasons for hip / elbow dysplasia from environmental issues such as weight, over exercise or poor diet. One last thing. as there is still far more data for hips than for elbows the system works much better for hips. we discussed how elbows are starting to be more of an issue than hips. he agreed that the kc data was starting to indicate this. I asked whether a more detailed elbow grading system wouldnt be more helpful than the 0,1,2,3 grades. he said that he thouht it would be more informative but this isnt a british sytem and we have adopted a more widely used scheme. europe has a, b,c, hip scores. I hope this is of interest to people. i certainly found this chat fascinating and reassuring when we think of all those affected by these worrying conditions.
Re: Estimated Breed Values That is very interesting, and makes a lot of sense. Thanks for explaining it so well.
Re: Estimated Breed Values That's fantastic. Thank you so much. Incredibly useful info. The link is here: http://www.thekennelclub.org.uk/services/public/mateselect/ebv/Default.aspx Here is the result for Naiken Kris (which I randomly looked up, but thought it might be useful to illustrate your great explanation NK EBV by JulieTandCharlie, on Flickr Brill. O - off to look up more!
Re: Estimated Breed Values Thanks so much julie. Im on my ruddy tablet and that was beyond my it capability. from naiken kris, I can see hes a well used dog of good repute as he must have a good number of offspring who themsekves have been hip scored. look how his hip score is more trusted than his elbow. Simply because more dogs are hip scored than are hip and elbow scored. The figure though has a very high confidence. can you copy two dogs for me julie? Brackenbird minnow and kupros master mariner? Only if you have a mo. thanks
Re: Estimated Breed Values For a constantly challenged statistical brain like mine.....remind me..... The higher the confidence interval, the better. Over 96% is pretty significant??
Re: Estimated Breed Values Brackenbird minnow is our eevees dad. ftch in a trendy colour and 13 years old! so well used and progeny tested. But never elbow scored due to his age, yet his elbows rank better than his hips which total 6.
Re: Estimated Breed Values [quote author=drjs@5 link=topic=4859.msg60519#msg60519 date=1394665138] For a constantly challenged statistical brain like mine.....remind me..... The higher the confidence interval, the better. Over 96% is pretty significant?? [/quote] Yes, although it should also give a range. So you should be able to say (eg) 96% confidence it's +/- x% from point Z. But the range is illustrated diagramatically. So on Naiken Kris, it looks to be 96% confidence the hip result is between -40 and -50 (about).
Re: Estimated Breed Values [quote author=hd link=topic=4859.msg60523#msg60523 date=1394665412] yet his elbows rank better than his hips which total 6. [/quote] They might be better - the range is big, and confidence lower. Thank you so much for this. It is hugely interesting. Got to go to bed rather than spend the next 4 hours working out the cummulative result for the whole of Charlie's pedigree. Which is pointless anyway, because I know what his hips and elbows are like from the x-rays. But love it. Data, data, data.
Re: Estimated Breed Values Well that's me getting nothing else done today... What a sensational tool. Love, love, love it. Thanks for your terrific post, hd!!!!!!!!!!!!
Re: Estimated Breed Values From what he explained to me, there would be very few dogs with a better confidence rate than 96% and therefore youd be statistically very unlucky if your puppy with him as the sire had hips only just better than breed average if it was created by genetics. The reason I mentioned kupros master mariner is that you can see the effect his untested elbows have obviously had on his progeny. That is one prolific dog too so it helps me to see how elbow problems have crept their way through our breed.
Re: Estimated Breed Values Julie, youre expecting too high confidence levels. youve got to look at the two dogs I posted progeny and see the number of non 0 elbows in their offspring. Youve got to take this tool in the way its meant - an overall way of isolating whether a dogs score is more statistically likely tobecaused by genetics or nurture. Not a way of predicting what one pup's score will be. Theres always going to be dogs that fall outside the expected range, youre just playing averages with no guarantees. Its a way of helping us to rid us of these awful ailments, not to provide us with 100% statistical confidence of a dogs reliabilty. Plus as I say, youll never get such levels of statistical confidence for a bitch so nothing can be cast iron.
Re: Estimated Breed Values [quote author=JulieT link=topic=4859.msg60527#msg60527 date=1394665668] [quote author=hd link=topic=4859.msg60523#msg60523 date=1394665412] yet his elbows rank better than his hips which total 6. [/quote] They might be better - the range is big, and confidence lower. Thank you so much for this. It is hugely interesting. Got to go to bed rather than spend the next 4 hours working out the cummulative result for the whole of Charlie's pedigree. Which is pointless anyway, because I know what his hips and elbows are like from the x-rays. But love it. Data, data, data. [/quote] I meant rank better on the ebv, he could have pants elbows but if he does it isnt going down to his offspring in the way that something from kupros master mariners obviously have.
Re: Estimated Breed Values [quote author=hd link=topic=4859.msg60531#msg60531 date=1394667185] Julie, youre expecting too high confidence levels. [/quote] No, no, I just meant statistically on the EBV. I'm not expecting anything. The range on blackbird touches -20, which is where the hips are (with a tighter range). I should probably have said yes, much more likely to be better.
