MSTN (myostatin)

Crossing two of them

Where the stamina went

In real horses this locus trades sprint against stamina. This mod has settled that there is no stamina resource (a settled decision), so there is nothing for the T side to buy — and a gene where one allele is simply better than the other is not a choice, it is a chore. So stamina is paid out in the nearest thing the game does have: max health. A stayer is the horse that keeps going, and in Minecraft terms that is the horse with more hearts.

Its gene carrot

Gene key

horsegenetics.mstn

Priority

80

Alleles

C T

Outcomes

sprinter, middle-distance, stayer

Coat effect

none — every outcome is a wild type

Default allele

T

Wild population

p(C) = 0.35 — both ends are common

Founder draws

1 × nextFloat()

Catalogue entries

1 — it never changes what a horse looks like

The combination table

Codominant, and therefore additive

Neither allele hides the other. Each C copy is worth 0.020 movement speed and costs 2.0 health, and the contribution just counts copies — which is all “incompletely dominant” has ever meant. No special case anywhere.

What myostatin does

MSTN — also called GDF8 — is one of the best-evidenced performance loci in the horse. It encodes myostatin, a transforming-growth-factor-β superfamily protein that acts as a molecular brake on skeletal-muscle growth: it limits myoblast proliferation and differentiation and constrains fibre growth. Less myostatin signalling generally means more capacity for muscularity, though what that looks like depends heavily on the particular mutation and species. PMC2913906

The equine gene is on equine chromosome 18 (ECA18), with three exons and two introns, spanning roughly 6.2 kb on the reverse strand around 66.49–66.50 Mb in EquCab2. Myostatin is made as a precursor and processed into an N-terminal propeptide and an active C-terminal signalling molecule.

So — unlike Belgian Blue cattle or homozygous myostatin-mutant whippets — ordinary performance-horse “speed gene” genotypes do not create a dramatic double-muscling syndrome. The equine variant is subtler: it shifts myostatin expression, and with it muscle architecture and exercise phenotype.

The two variants that matter

The intron-1 C/T marker

The widely marketed Thoroughbred test targets a T-to-C SNP in intron 1, in legacy notation g.66493737C>T, giving C/C, C/T and T/T results. It is intronic and changes no amino acid. The initial association studies found it an exceptionally strong predictor of optimal racing distance.

C/C horses were common in the short-distance cohort and absent from the longer-distance cohort; the National Hunt horses in that sample likewise had none. These are tendencies, not destiny — a C/C horse can be a poor sprinter and a T/T horse can win a short race. Hill et al., PLoS ONE 5:e8645

The 227-bp promoter SINE insertion

The key advance is that the intron-1 marker is usually in linkage disequilibrium with a 227-base-pair SINE insertion in the MSTN promoter. A SINE is a short interspersed nuclear element — mobile, repetitive DNA that landed near the gene’s regulatory machinery. It is now the best-supported functional variant:

• It reduces MSTN production in cell experiments by about 4.5-fold.
• Horses carrying it have lower circulating myostatin.
• The effect is dosage-related: two copies lowest, one intermediate, none highest.
• In Thoroughbreds, insertion homozygotes have a higher proportion of fast type-IIX fibres and fewer slow type-I fibres than wild type.

That changes how the old C/T test should be read. The SNP is an excellent predictive marker in populations where it tracks the insertion; the promoter SINE is more likely the causal change. They are not interchangeable across every breed or family, because linkage patterns differ, and for cross-breed work testing the SINE directly is cleaner. Sci. Rep. s41598-025-22472-7

Everything else at the locus

Other promoter and intronic SNPs are documented — g.26T>C and g.156T>C have different frequencies across heavy, mesomorphic and light horse populations, and one sits in a TATA-box-like motif. Plausible regulatory candidates; nothing like the SINE’s causal evidence. There are also multiple MSTN haplotypes rather than one universal fast allele and one slow allele, which is why simple cross-breed interpretation is risky. PMID 25160752

Build, muscle and size

The clearest body relationship is not height but muscle mass relative to frame. In Thoroughbreds measured in their two-year-old seasons: C/C averaged 2.94 kg per cm of wither height, C/T 2.88, T/T 2.83 — and in males, C/C averaged about 6.7% greater mass per unit height than T/T.

So the C/insertion end is more compact and power-oriented with a fast-fibre bias, and the T/no-insertion end is relatively lighter with more stamina-oriented muscle physiology. None of which means MSTN determines a good shoulder, a broad hip, a short back or big bone. Those are highly polygenic and shaped by breed selection, sex, maturity, nutrition and training.

MSTN is not a primary height gene. The strongest stature signals are elsewhere, notably the ECA3 region near LCORL/NCAPG, which in a large German Warmblood study explained far more withers-height variation than any putative MSTN effect. PMC11177368 “C/C makes a horse smaller” and “T/T makes a horse taller” are both unsupported. Older breed-comparison work found promoter allele frequencies differing between heavy and light morphological groups, and the authors explicitly noted that linkage to another ECA18 locus, or population structure, could explain it.

Speed, stamina, jumping

Flat-race speed is the strongest evidence base. Lower myostatin from the promoter insertion is consistent with a fast-twitch, power-oriented phenotype; in sprint conditions a horse benefits from rapid force development and glycolytic capacity, so the insertion-homozygous state associates with short-distance aptitude, heterozygotes are intermediate, and no-insertion horses more often suit sustained racing.

The nuance that matters: MSTN predicts distance preference far more convincingly than it predicts absolute maximum speed, earnings, soundness, temperament, start quality or the ability to beat a particular rival. It describes an engine’s tuning, not whether the car wins.

Endurance. The T/no-SINE profile associates with longer race-distance aptitude, not with success in competitive endurance riding, which also turns on thermoregulation, metabolic efficiency, cardiovascular capacity, hydration, management, feet, musculoskeletal durability, rider strategy and many other genes. The original study included Egyptian Arabians in which T/T was very common, consistent with selection for endurance-type physiology — and not proof that MSTN alone makes an endurance champion.

Jumping. There is no robust evidence that MSTN is a major jumping gene. Jumping depends on canter quality and adjustability, coordination and balance and rider partnership, fore- and hindlimb conformation and spinal mechanics, tendon and ligament integrity, and on careful take-off judgement, agility, willingness and technique as much as on power. A fast-twitch bias plausibly helps the power component — a mechanistic hypothesis, not a selection rule. Reviews list MSTN among broad athletic-performance genes; that is not proof an MSTN test predicts jumping talent. The mod gives MSTN no jump effect, which matches.

Health and injury

In Thoroughbreds, the intron-1 marker linked to the promoter SINE has been associated with susceptibility to carpal fracture — plausibly through the combination of greater body mass, higher speed and repeated high loading of immature racehorse limbs.

Read that carefully. It is an association, not a diagnosis; it does not mean the genotype causes fragile bones; and race surface, training load, age, growth rate, shoeing, conformation, prior injury and management are all major contributors. Genotype does not substitute for sound conditioning and veterinary monitoring.

Not established: that common MSTN performance variants directly cause a general inherited muscle disease, PSSM1, HYPP, HERDA or any other named disorder; a universal disease-prone phenotype; a predictable joint, tendon or ligament condition outside specific study contexts; or a generally healthier or less healthy horse. Recent work also shows circulating myostatin is fairly consistent within an individual over time, while stating explicitly that it is unresolved whether differences within the same SINE genotype group predict performance, muscle phenotype or fracture risk.

Coat

No known role in coat colour, white patterning, grey, dilution, mane and tail colour or coat quality. MSTN is on ECA18 and regulates muscle; colour lives at MC1R, ASIP, STX17, KIT, MITF, EDNRB, PMEL and their neighbours. If an MSTN type “often appears” in a particular colour, that is breed history, sire-line popularity or selection practice.

Inheritance

Autosomal, on chromosome 18. Each horse has 0, 1 or 2 copies of the insertion; two heterozygous parents give 25% / 50% / 25%. The molecular inheritance is codominant and additive in the sense that insertion dosage tracks average myostatin level. For the C/T marker the same 0/1/2 logic holds, but since it is usually a linked marker it is best not described as simply dominant or recessive.

And the phenotype is not Mendelian in the “genotype equals outcome” sense. The variants segregate by Mendel; race-distance aptitude, muscle composition, build and injury are complex traits, and the three genotype classes express a graded pattern with large overlap between individual horses. The mod, having no way to express overlap, renders the tendency as a fixed number — which is the general shape of the compromise this whole model makes.

Breeds and frequency

Thoroughbreds are the most-studied breed and the SINE is common. One 14-breed haplotype study put SINE frequency at about 0.73 in Thoroughbreds; another reported 0.51 in its sample — published frequency moves with population, country, sampling, and whether the horses are selected for a racing subtype. The original performance study showed how sharply class selects: the C allele was 0.72 in its short-distance elite cohort against 0.36 in the longer-distance cohort. That is not a population estimate; it is evidence of how strongly training and breeding choices enrich a performance subgroup.

Quarter Horses are the other high-prevalence breed, fitting a history of selection for high-power short-distance racing — and published samples differ markedly: about 0.48 SINE frequency in one multi-breed study, 0.81 in a smaller Italian sample, and in an earlier sample of 35 Quarter Horses a C-marker frequency of 0.90 with C/C in 83% of horses. Common to very common, and not a fixed allele.

Other breeds. The intron-1 C allele is more widespread than the functional SINE: across 301 horses from 14 breeds it was present in 12 of 14 breeds at an aggregate ~0.27, while the promoter SINE was detected in only five breeds and was common only in Thoroughbreds and Quarter Horses.

In Egyptian Arabians from the original study T/T was 90%. In Miyako horses, a Japanese native breed, the intron-1 variant sat at minor-allele frequency 0.04 and the SINE was absent. PMC12995546 In Mongolian native horses the C allele ran around 0.11 overall (0.06–0.15 by population) and the insertion was not detected among 381 genotyped horses — and those researchers found no association between the SNP and seven measured conformation traits, which is a good reminder not to export Thoroughbred conclusions to unrelated breeds. Anim. Behav. Genet. 54(2)

The mod uses p(C) = 0.35, deliberately common at both ends so a player finds both a sprinter and a stayer in their first handful of caught horses. Against the real numbers that is a world-average sitting between the Thoroughbred/Quarter Horse end (0.48 to 0.9) and the many breeds where the insertion is absent altogether — which is arguably the most defensible thing the mod could do with a locus whose real frequency is almost entirely a statement about breed rather than about horses.

What is settled, and what is not

Strongly supported. MSTN is on ECA18 and encodes a negative regulator of muscle development; the 227-bp promoter SINE reduces expression and is the best-supported functional explanation for the speed-gene effect; in Thoroughbreds the genotype is strongly associated with optimal race distance in the sprint / intermediate / stayer pattern; the effect is tied to measurable changes in circulating myostatin and muscle-fibre composition; and the SINE is common in Thoroughbreds and Quarter Horses and uncommon or absent in many other sampled breeds.

Plausible but incomplete. That the fibre effects influence burst power, acceleration and recovery beyond simple race distance; that the locus contributes to body mass relative to height in specific sex, breed and training contexts; and that the carpal-fracture association reflects a real interaction between speed, mass, bone loading and immature racing schedules.

Not established, or overclaimed. That MSTN determines coat colour; that it reliably predicts height, broad conformation, temperament, trainability or overall athletic talent; that it is a validated jumping test; that it identifies a best breeding match without pedigree diversity, conformation, soundness and the rest of the genome; that it proves a horse healthy or unhealthy; or that a commercial C/T result alone is a causal diagnosis.

The defensible wording: MSTN genotype helps describe an inherited tendency toward a more sprint-power or more stamina-oriented muscle phenotype, especially in Thoroughbred-derived racing populations. It does not determine athletic worth, soundness, coat, size or jumping ability.