@2017 by Performance Genetics LLC

Speed is not Class

February 1, 2012

Authored by -



A lot has been written in the last week in regards to the concept of the ‘speed gene’ and exactly how and when this variant (change or mutation in the DNA code) found itself in the thoroughbred population. Firstly, it must be clearly stated that we don’t take any issue with the science of the paper in terms of how the research was conducted and completed. It was good work. The paper also shows how modern genetics can be useful in solving intriguing questions regarding the history of the thoroughbred, which is something that we believe is well overdue, so at a certain level we are really excited that the paper received such widespread coverage. In addition, the paper also demonstrated how racing environments can select against certain traits, but when the environment changes, the population can adapt their genetics to respond to those changes. That being said, we do have some questions as to the conclusions that were made and how the paper has been subsequently interpreted by the media.


As far as the "C" allele variation found in the MSTN (myostatin) gene is concerned, the paper makes a interesting claim as to the exact source of this variant in the so-called 'speed gene.' The press release (moreso than the paper) states that the 'speed gene' entered the Thoroughbred from a single contributor, and that this contributor was a British mare born about 300 years ago. Leaving aside that the gene itself exists in all equine breeds and didn't insert itself magically into a population as some media outlets reported, the paper itself doesn't entirely  substantiate the conclusion at which it arrives. 


While recent analysis of equine mitochondrial DNA variation has demonstrated that a large proportion of Thoroughbred foundation mares were from native British stock, that is a separate issue to the nuclear based variant within the MSTN gene. After all, in terms of genetics, it only takes one popular stallion or influential mare to insert a genetic variation into the population and it takes a human lifetime (or generations) to get it out! The paper starts the line of inquiry off in terms of identifying the source of the "C" allele by establishing two horse populations with relatively high frequencies of C-alleles on the MSTN gene, the Fulani, (with 33% of that population carrying the "C" allele) and the Shetland (with 50% of the population carrying the "C" allele). The link between the Fulani and Thoroughbreds is clear as the Barb horse, originating in North Africa –  as did the Fulani – is known to have contributed to the foundation of the Thoroughbred. While recognizing the presence of the "C" allele in the Fulani the scientists themselves state that "The absence of haplotype data for Fulani (or any other representative North-African population) precludes rejection of the possibility that the C-allele originated from a stallion or mare of Barb origin." Essentially, even though approximately 33% of the Fulani breed may carry the "C" allele, even  though the Fulani is a Barb breed,  and even though the Barb has been a very important contributor to the Thoroughbred, because they didn't have data on the Barb, they discounted the possibility of it being a potential  source either along with, or over the Shetland (whose own qualification, according the paper, is no stronger than being associated with the same geographic region as the British native running horse, the now-extinct Galloway).


The paper further states "...The historic samples were related largely via the Darley Arabian sire line to which 95% of all living Thoroughbreds can be traced in their paternal lineage. As neither the Byerley Turk nor the Godolphin Arabian sire lines were sampled in the historic data, we cannot preclude them as a source for the C-allele. However, as the Darley Arabian has had a disproportionate influence on living male lineages, and hence is overrepresented in this sample of prized stallions, it is more likely that the C-allele entered the Thoroughbred gene pool via a maternal lineage...." Again, the researchers discount the possibility that the "C" allele may have come from either the Byerley Turk or the Godolphin Arabian (also known as the Goldolphin Barb, and very possibly of North African origin) only on the basis that they were not able to get historical samples from horses from that sire line. As well founded as our concern is, we would be as guilty of the authors of the paper in speculating where and when the variant within the MSTN gene that is reference actually entered the population, and when and by whom it truly proliferated (we ourselves suspect, although without concrete evidence, the stallion Phalaris), so it is probably best left that in regards to how the “C” variant entered the population –  ‘alternate possibilities to a sole native British mare do exist’.


This does however bring us to a significantly more important discussion on the exact term ‘speed gene’ and the concept that a single variation within a single gene determines optimal racing distance.

From the time the 'Speed Gene' was first announced to the thoroughbred community, in a commercial sense, and the subsequent launch of the the ‘Speed Gene test’, media releases and papers coming from the scientists behind the test have never made it clear that when they use the term "speed" they are referring to aptitude - in the sense of speed v stamina - not class, wherein "speed" would refer to the ability to cover a given distance faster than a contemporary. For example, Yeats showed greater speed over 2½ miles than his opposition in his Ascot Gold Cups, even though some may have had more "speed" in that they could have had greater predilection to sprinting. One suspects that the blurring of the line between "speed" in terms of aptitude (sprinters vs. stayers/routers) and "speed" (in terms of assumed class) may not be an accidental one.


The concept that a single variant within a single gene is a determinant of optimal racing distance is also something that we have disagreement with. It is worth noting that at the beginning of the paper, the authors rightly state...."In humans more than 200 genes have been reported to be associated with fitness-related health and exercise traits, and it is likely that racing performance in the Thoroughbred is also polygenic and is influenced by genes that contribute to the wide range of anatomical, metabolic and physiological adaptations that enable elite-racing performance." Again, we want to be clear that we are talking about optimal racing distance or aptitude here, not discriminations of class.  In terms of the former, however, we don’t believe it is completely accurate to suggest to the scientific and wider thoroughbred community that racing performance in terms of class is polygenic (meaning that multiple genes are contributing) but in terms of racing aptitude, it is best determined by one single variant (change in DNA code) within one single gene.  While the MSTN gene (located on Chromosome 18) is indeed important in terms of both racing class and aptitude, it is certainly not the only gene that contains variations within it that help determine the optimal racing distance for a horse located on ECA18, ECA18 isn't the only chromosome with variants that help determine optimal racing distance and the SNP that Equinome established isn't the only variation in NSTN that influences distance.


Prior to, and since, its commercial launch Equinome has spent a lot of time, money and research in establishing the particular variant ( g.66493737C>T) within the MSTN gene.  However, around the same time as Equinome's first paper on the variants within MSTN came out, Dr. Matthew Binns established two other SNPs in ECA 18 that influenced optimal racing distance in North American racehorses. In his genome-wide study Binns identified BIEC2-417274 (located at g.65868604 bp on ECA18) and BIEC2-417495 (located at g.67186093 bp on ECA18), both either side of Equinome's MSTN variant located at g.66493737, as determinants of optimal racing distance in the North American thoroughbred.Subsequent to this, Performance Genetics' own research conducted earlier this year, and using the recent released Equine70KSNP Chip, has identified three SNP's, one located on the MSTN gene, one located 5 million base pairs downstream of the MSTN gene on Chromosome 18, and another SNP on a different Chromosome altogether, that also have an influence on optimal racing distance in thoroughbreds. Thus, between these three companies, we are talking about no less than six variants that have an influence on optimal racing distance alone. Interestingly Equinome also published that a SINE insertion within MSTN also had an influence on optimal racing distance but they did not further investigate its influence.


From a practical viewpoint, much of the paper had little application to modern thoroughbreds. It is, as we said earlier, an interesting historical piece on genetics within the thoroughbred, showing how a population can come under selection pressure for desirous traits, depending on the environment that it is placed within. In terms of application of an optimal distance test, while it is useful to know if your horse is an early maturing/sprinting type as opposed to a later maturing/distance type, and make management decisions based on this, we have seen ourselves horses that are genetically predisposed to being sprinters that depending on the circumstances can run further than their genetics would immediately suggest. This occurs particularly frequently late in the two-year-old and early in the three-year-old stages of a horse’s career, when many of the horses that should excel at a distance are not mentally mature enough to do so, and the class or seasoning of a sprinter, or a slow pace, allows them to be competative beyond their optimum distance.


What of course is more useful, is a test for class. Late last year Performance Genetics launched Sales Select, a genetic, cardiovascular and splenic capacity test. The genetic component, which was developed by comparing groups of elite and non-elite horses, uses multiple SNP’s to ascertain both the optimal racing distance and class potential for a racehorse.

Please reload

Featured Posts

Mitochondrial DNA haplotypes and mating outcomes

September 3, 2018

Please reload

Recent Posts
Please reload