CSV statistic

[elf]

Hi, here is how it's processed, for each dog if there is owner informations I put for the dog as living place the country the owner entered. I guess this works most of the time, but as we see not 100% (e.g. if the owner entered its contry of origin and lives outside it, or if there is co-ownership).

The best way for changes would be to modify (PM to Admin ?) your informations http://www.wolfdog.org/fra/dbase/o5142.html ; you can also add the city and your dog will be showed (next update) in "Find dogs living in city:".

[Nebulosa]

Thank's very much, Elf !

[elf]

Originally Posted by Nebulosa

Thank's very much, Elf !

I added for "Find dogs born in year:" the mean COI (on 5 and 8 generations) for the chosen year. This gives since 1990:

Mean COI on 8 generations:



Mean COI on 5 generations:



So if the former rule was to try to be less than 10% on 5 generations, let's roll the new rule, less than 5% on 5 generations.

[saschia]

Nice work Elf.

Would it be possible to have sorting option (or filtering), for for example the country of origin, HD, ED, bonitation code etc?

[elf]

Hello Saschia, I added a filter by country for:

"Find dogs born in year:"
"Find dogs with HD:"
"Find dogs with ED:"
"Find dogs with PRA:"
"Enter expression to find in bonitation:"
"Enter expression to find in training:"

If one want no filter by country just select "All countries".

[saschia]

So fast ;o) thanks!

[elf]

Originally Posted by saschia

So fast ;o) thanks!

Yeah I was in the process of adding it .

Added also average AVK (in "Find dogs born in year:" and "Find breeder"). This gives since 1990:

Average AVK on 8 generations:



Average AVK on 5 generations:

[Mikael]

Originally Posted by elf

let's roll the new rule, less than 5% on 5 generations.

YES !!!

And one thing more, even if I do not wont you to work your self to death is it possible to add whitch dogs are not alive ???

And I just wont to add again, Excellent work Dr Anthony The breed loves you Auuuuuuuu- After all we wont happy wolf not crying wolf

Best regards / Mikael

[saschia]

Originally Posted by Mikael

After all we wont happy wolf not crying wolf

I think Edit might object slightly

[elf]

is it possible to add whitch dogs are not alive ???

I don't have this information, I guess it will be in the next database, maybe Admin would be able tell us.

[saschia]

I think this might be a problem, as not all deaths are accounted for and then you wouldn't know which dog is alive and which is just not reported. Unless you took the age of say 18 years as suspected dead unless proven alive.

[Nebulosa]

Originally Posted by elf

I don't have this information, I guess it will be in the next database, maybe Admin would be able tell us.

Uhm.. all death dogs are marked with a red X at the side of their name.
But I wonder if will be possible to pass this information also for the statistics in the easy way.

[buidelwolf]

Hello Antony,

Really a fantastic tool as I already wrote. I've already made several interseting studies with it.

I added for "Find dogs born in year:" the mean COI (on 5 and 8 generations) for the chosen year. This gives since 1990:

Many dogs, born in the last years, seem to have increasingly low COI, measured on five (or eight) generations. This seems to be a positive development, but .... those appearances are a little bit deceiving, as the real coi reveals in fact on all generations, not only on 5 or even 8 generations.

What worries me a little bit is that I discovered that almost all dogs, even in the last years, have a COI, measured on ALL generations, of about 20% or higher. You rarely see one with a COI less than 17% measured on ALL generations. I wonder if it could be possible to measure the COI of all dogs born in year X measured on all generations? It might be very interesting to discover whether the COI measured on ALL generations in recent years is reduced and to what extent.

Anthony, thanks again. Great job!

Great tool! Well done!

Can you add a search, who find out where the most dogs in a country come original from?
For example in France: are the most in foreign countrys bought dogs from Czech Republik or Slovakia or maybe Italy?
That will be very interesting.

[saschia]

Originally Posted by buidelwolf

Many dogs, born in the last years, seem to have increasingly low COI, measured on five (or eight) generations. This seems to be a positive development, but .... those appearances are a little bit deceiving, as the real coi reveals in fact on all generations, not only on 5 or even 8 generations.

Yes, this is a problem in some way. We are working with a closed group of animals, which therefore are all related to each other to some degree. On the other hand, that's (if you take all generations) true for all breeds and actually all species as well. So one thing is, how far in the number of generations, is it sensible to go.

On the other hand, although the animals from for example one litter are very related, they nevertheless have different genes, different expression levels of the same genes etc. Plus there are always changes in the genome in the form of mutations, deletions, duplications etc. So after some number of generations these changes may have larger effect than the relatedness of the animals. So that's another reason why taking into account all generations is not sensible.

So, in breeding it is good to take into account the COI and AVK coefficients, but not use them as the only leading tool - because you can get them closer to ideal values by just breeding a lot with young animals therefore increasing the number of generation in certain time interval, while it is important to select which animals to use according to their health, character and exterior.

[elf]

Uhm.. all death dogs are marked with a red X at the side of their name. But I wonder if will be possible to pass this information also for the statistics in the easy way.

Yeah difficult, I would need an access to the database to add it.

What worries me a little bit is that I discovered that almost all dogs, even in the last years, have a COI, measured on ALL generations, of about 20% or higher.

There are even COI of 40%, 50%, 60% on less than 8 generations, which is really crazy. AVK is also very informative.

I wonder if it could be possible to measure the COI of all dogs born in year X measured on all generations? It might be very interesting to discover whether the COI measured on ALL generations in recent years is reduced and to what extent.

I actually can't, the code is too slow, I need to switch to more efficient calculation method.

Can you add a search, who find out where the most dogs in a country come original from?
For example in France: are the most in foreign countrys bought dogs from Czech Republik or Slovakia or maybe Italy?

I added this in "Find dogs living in country:" ,you have now a new field "Dogs from:".

Plus there are always changes in the genome in the form of mutations, deletions, duplications etc. So after some number of generations these changes may have larger effect than the relatedness of the animals.

And epigenetic modifications are transmissibles... but I don't know the rate of apparition/transmission.

For the future, I think dog's mean kinship coefficient would be one of the very important coefficient to take into account when one choose how many time to reproduce a dog. I cannot provide this coefficient in real time in the tool because it would require long time of calculation for each dog, but I maybe could do it offline, then add it in the DB.

[elf]

More than 10000 connexions, glad people are using it, hope it helps.

Added in "Wolfblood" GSD composing the dog, no reason to forget them . Exemple for Furcas: Furcas Wolfblood

Following graphs representing size repartition (from bonitation code) Female and Male (%/cm):

Female:



Male:




Eventually I calculated mean kinship coefficient for all dogs born after 1995:

"A conservation biologist would consider the individual with the lowest mean kinship to be the most genetically valuable in terms of maintaining diversity in the population, and would try to favor that individual in a breeding program."

I decided not to publish them as it could represent a kind of ranking that may be used in a wrong way. Anyway if at some point it may be needed in some programs keep in mind that it does exist.

Cheers and happy new year to all, two and four legs !

[Mikael]

Thanks this was very good info

Best regards / Mikael

[Rona]

Elf, is your tool connected to the wolfdog.org database? If not, is it possible to update the data from time to time (every 6 months or every year), or would it be too complicated? I mean, lots of pups were born recently, many dogs were bonitated and tested, many gained breeding rights and this might have changed a bit the overall statistics...

On the other hand, I think it would be interesting to keep the consecutive versions intact in some archives for further comparisons, to see the directions of the breed developement...

[elf]

Unfortunately the tool in not connected to the wolfdog database, I can update from time to time (last DB update in October) but it's a bit complicated and I don't success to motivate myself to do this often . The best way ahead would be to make this tool a part of wolfdog.org and use directly its database, I can provide help, it's up to the Admin.