Coefficient of Inbreeding

[Whatif]

Advertisement - Members don't see this ad

Hello,

I hope someone would help me find out what I'm doing wrong calculating the inbreeding coefficient for this dog. Here is a link to the pedigree:

http://www.pedigreedatabase.com/german_shepherd_dog/pedigree/5/printer/450338.html

Kriemhilde von Nassau is linebreed on:

* 4 - 5,5........................................... in V Graf Eberhard vom Hohen Esp
* 4,4 - 3........................................... in VA1 Tell von der Kriminalpolizei
* 5,5 - 4........................................... in SIEGER 1908 Luchs vom Kalsmunt Wetzlar
* 5,5 - 4........................................... in Herta von der Kriminalpolizei
* 3,5 - 4........................................... in V Munko von Boll
* 3,5 - 4........................................... in Hella von Boll
* 4 - 5............................................. in Lori von Brenztal
* 4 - 5............................................. in Achim von Tautenburg
* 4 - 5............................................. in Minka von Boll
* 3 - 2............................................. in VA (NL) Jung Tell von der Kriminalpolizei
* 4 - 5............................................. in Fanny von der Kriminalpolizei
* 4 - 3............................................. in Gerta von Boll


When I calculated the COI for this dog, I got 11.9141%. I'm using a web based COI calculator and a software to compare my result to them. The web based calculator is giving me 14.84375% while the software is giving me 12.2070%. I doubled checked everything and still get the same results. I only inputted the dogs in the first five generations.

Your help will be much appreciated,
Mike

 

[nyanko]

Hi!

Not sure if this is really the right place for this type of question or not (really kind of not ), but it would help if you said what methods you were using to calculate inbreeding coefficient, and what software you were using too. You can get a lot of different results for the same pedigree depending on many different factors, but honestly the results you got are not all that far apart in the first place..

edit: It's also going to be pretty wildly inconsistent given that it only goes back that number of generations. Not sure it's really a useful metric in this case.

 

[squibby]

Good god, I don't understand any of that.

 

[nyanko]

At its most basic level an inbreeding coefficient is essentially a measure of the parents' relatedness to each other. Numerically it represents an estimate of the proportion of the genes present in the individual expected to not only be homozygous (identical by state), but with both alleles (maternal and paternal) originally derived from the same ancestral allele (identical by descent).

Wikipedia talks about identity by type pretty succintly.

 

[Whatif]

Thank you for taking the time to reply!

Couldn't find a better place to place this question......I'm using Wright’s equation. The web based COI calculator and software won't reply to my inquiry so I don't know the method they use. But, shouldn't we all get the same result despite the method? Which method is the most accurate?

I have only inputted the dogs in the first five generations in the two calculators and used the first five generations in my own calculation.

 

[nyanko]

I'm using Wright’s equation. The web based COI calculator and software won't reply to my inquiry so I don't know the method they use. But, shouldn't we all get the same result despite the method? Which method is the most accurate?

The short answer is no, the different methods of calculating inbreeding coefficient will not give the exact same answer, as it is an estimate. There's really no way of knowing which is the most "accurate" - they're probably not statistically significantly different from each other anyway given the effective population size.

What do you need to use this calculation for?

 

[Whatif]

I'm trying to write a program to calculate the COI for further analysis. Do you know of any sources that explains the tabular method? or any method of calcualting the COI other than the path method? I searched online without success.

 

[Raimes]

What do you need to know about the tabular method? Just how to do it? Or explaining it's reasoning behind it?

 

[Whatif]

How to do the tabular method and the reason behind it.

 

[nyanko]

Maybe you should learn a little bit more about pedigree analysis before you attempt to write a program about it?

 

[Whatif]

Well, I have been using Wright's equation and calculated over 40 cases, many of them are complex. I got stuck with 3 cases, which I have different result than the COI programs and I'm seeking help. I know not too many people would be willing to do these types of calculations but was just hoping if at least someone one could direct me on where to seek help.

 

[david594]

Is this like a homework assignment or something?

Just post up the instructions, data set, and a copy of your code and we'll see what we can do for you....

 

[alliecat44]

SDN is not for homework help. Closing thread.