This morning you may see that we have extended the computations of the
Coefficient of Inbreeding
and
Coefficient of Relatedness
to 10 generations. These coefficients were previously calculated over 8 and 6 generations respectively and were done to that generation previously due to computational restrictions, i.e, the speed of the computers and the frequency of TrueNicks requests being run had an effect on report delivery time.
We have been able to extend the computation out to 10 generations thanks to the programmers at TJCIS, and it is also where a nice balance is found between the calculations of these coefficients and pedigree completeness. As we go further back into a pedigree, the completeness of the page starts to reduce as names are either founder horses, where no ancestor is known, or names where ancestry cannot be reasonably verified. This incompleteness has an effect on the calculations for inbreeding and relatedness as they then start be underestimated, especially where the incompleteness is skewed to one side of the pedigree over another.
As a reminder, we first placed these coefficients back in 2013 - you can see the blog post for that
here.
The coefficient of inbreeding (COI) was introduced by Sewell Wright back in 1922 to express the expected percentage of homozygosity (closeness to being identical) arising in a given mating. It can also be viewed as the average chance that any one gene pair is homozygous due to inheritance from a common ancestor. A low inbreeding coefficient means a low level of inbreeding, at least in paper terms. The vast majority of racehorses have an inbreeding coefficient of less than 5%. Inbreeding coefficients over 5% are unusual, and over 10% are very rare.
Recently a paper by
Todd and colleagues
pointed to higher levels of inbreeding having a negative impact on racehorse performance. Their analysis of data from 135,572 Thoroughbred horses in Australia revealed a strong negative relationship between Wrightâ€™s inbreeding coefficient and five measures of racing performance that encompass a range of factors that contribute to exercise performance. These included two measures that are based on the assumption that more successful individuals earn more prizemoney: cumulative prizemoney earnings and prizemoney earnings per start. They also included two measures of constitutional soundness: total number of race starts and career length and accounted for consistency of performance with the measure winning strike rate. The authors suggested that the negative relationship between inbreeding and performance can be explained by a genetic load of partially deleterious alleles still being carried by the population.
The Coefficient of Relatedness (COR) answers a slightly different question. The coefficient of relatedness provides a way of objectively assessing the similarity of two pedigrees by giving a number that is a direct measure of shared ancestry. In this case, the two pedigrees that we use are the sire and dam of a hypothetical or named horse mating. This figure will vary from the COI because it is possible, indeed quite frequent that a sire or dam can be inbred themselves, thus contributors to the inbreeding of their offspring, but the level of inbreeding between the sire and dam can be low. That is, one parent can be intensely inbred while the other a relative outcross. There has been no studies on the COR and its relationship to performance.
In pushing these two metrics out to 10 generations, we have also added the Pedigree Completeness to 10 generations, as well as a count of the unique ancestors in a pedigree from the first to the 10th generation. This also sets the platform for us to implement the use of the Ancestral History Coefficient, which in the paper from Todd and colleagues above, had a positive influence on racetrack performance. With this latter metric in place, it will give us the platform to develop TrueNicks 2.0, a significantly more predictive algorithm for matings, which we have in the works to develop and launch in 2019.