Inbreeding is about doubling genes.

 

The more ancestry you have shared between the sire and dam, the more likely the foal is to get the same copy of a gene from its sire that it got from its dam.

 

This is important to racing ability since when a gene variant is doubled, its impact on the Horse will usually be intensified due to the effect no longer getting diluted by an alternative version of the gene.

 

This can be very good is the gene variant is good, but very bad if the gene variant is bad.

 

But why would a foal you bred get poisoned by the unwanted presence of bad doubled gene variants?  After all, you tried to inbreed to a good ancestor and that ancestor’s genes went through several generations before they got to your foal, generations of the fit reproducing and the weak being selected against that ought to have filtered out the bad genes.

 

But here’s the problem and here is the Achilles’ Heel of Inbreeding: The central reality of Thoroughbred Breeding is that the average foal has a sire that got the privilege of breeding by passing meaningful selection tests, while at the exact same time he’ll have a dam which got to breed without passing anything resembling the sort of selection tests that a sire would have.

 

Therefore each generation that the genes of the inbred to ancestor went through a female to get to the foal is a generation where there was no meaningful selection and consequently no meaningful filtering out of the bad genes.

 

Recently, I found strong evidence that this lack of female selection has a powerful real world effect on what kinds of inbreeding are more successful and what kinds are much less so.

 

How it started is that Adrian Parry sent me data on horses in his database that were born from ’90  to ’02 and which were inbred within 5 generations to a single ancestor.

 

This data was organized around which position that Common Ancestor was in on the Sireside and the Damside.  So that, for instance, you could see the % of SWs for cases where the Common Ancestor showed up in the Grandsire Position on the Sireside and 3rd Damsire Position on the Damside, etc. 

 

Also, in the data the Pedigree Positions were signified by Algebraic Notation where A1 and A2 are the sire and dam, B1 the Grandsire, etc.; as demonstrated in this Pedigree Chart of 2007 BC Classic Winner Curlin:

 

 

 

On account of my theory that Inbreeding would be compromised by the genes of the Inbred to Ancestor not getting filtered enough the times they went through females, I decided to check whether there was a Correlation between the number of Male Horses the Genes of the Inbred to Male Ancestor had to go through on the Damside to get to the inbred foal and the % of SWs for that kind of inbreeding.

 

For instance, if you inbred to the 3rd damsire of the potential foal, the genes of that Inbred to Ancestor would’ve went through 0 Males to get to the foal (as they would’ve just gone through the 3rd dam, the 2nd dam, and finally the 1st dam).

 

What I ended up finding was a strong correlation between the Number of Males a Damside Inbred to Ancestor had to go through to reach the foal and the % of SWs for that kind of Inbreeding:

 

 

What I found looking at the Spearman Rank Order Correlation between the % of SWs for one of the above categories of inbreeding and the Number of Males in the Chain leading to the inbred foal was incredibly strong r of .8932.  This means that 80% of the variation in the % of SWs from position to position is explained by the number of males in the chain leading from that position to the inbred foal.  ‘

 

(NOTE:  In contrast to the 80% Correlation between # of Chain Males and % of SWs, there’s only a 52% Correlation between the Number of Generations back a position is and its % of SWs; which shows that the number of Chain Males is the true driving force here and not overall Generational Distance.) 

 

Another way of looking at this is to note that on the Damside, the positions with 0 Chain Males between them and the foal got only 5.80% SWs in Adrian Parry’s database, that the positions with 1 Chain Male got 6.47% SWs, the ones with 2 Chain Males got 7.57%, and the one with 3 Chain Males got 8.73% SWs.

 

The most interesting thing to me though is that when I did the same thing with the cross of the Inbred to Ancestor found on the Sireside, nothing resembling such a strong Correlation was found:

 

 

Which is only a Spearman Rank Order Correlation of .0658 (which means that only a piddling 0.43% of the variation in % of SWs between positions are explained by the number of males in the chain leading from the inbred to ancestor to the foal).

 

Another way of looking at this is that on the Sireside, the positions with only 1 Chain Male got 7.38% SWs in Adrian ’s database, the positions with 2 Chain Males got 7.16%, the positions with 3 got 7.20%, and the position with 4 Chain Males got 7.33%.

 

To me the simplest and most natural explanation for these findings is that Dams are SO unselected as compared to Sires, both in terms of individual quality and pedigree, that whereas there’s a strong need to get the Genes of an Inbred to Ancestor filtered through as many males as you can when they come through the dam, a similarly strong need does not exist for the genes coming through the sire.

 

The lesson?  That if you want to inbreed within 5 generations, you should to try to inbreed to ancestors of your mare that had their genes go through as many male ancestors before reaching your mare as possible.

 

-Michael Ventura, December 14, 2007