What does the published literature report about genomic and genetic testing in horses for breeding decisions, covering: (1) inherited diseases that breeders test for, such as HERDA, HYPP and GBED, inc

Inherited diseases tested for breeding: causal variants, carrier frequencies, and mating guidance

Causal variants have been identified for fewer than 20% of suspected equine genetic traits (Direct; PMID: 38600096), but for those that have been molecularly defined, population genotyping reveals a strongly stratified carrier landscape. In a large survey of elite American Quarter Horse performance subgroups and age-matched controls, whole-breed allele frequencies were PSSM1 0.055, GBED 0.054, HERDA 0.021 and HYPP 0.008, while the related American Paint Horse population carried the lethal white foal syndrome (EDNRB) allele at 0.107 (Direct; PMID: 30623495). Within performance subgroups these values were several-fold higher: HYPP reached 0.299 in halter horses (56.4% of sampled halter horses able to transmit the allele), PSSM1 0.155 in halter horses (28.2% carriers), HERDA 0.142 in cutting horses (28.3% carriers) and GBED 0.132 in western pleasure horses, whereas racing and barrel racing subgroups were largely free of the alleles (Direct; PMID: 19119976). The paper's mating arithmetic is explicit: for recessive diseases, carrier-to-carrier matings produce affected foals 25% of the time, whereas for dominant conditions a single heterozygous parent passes the allele to 50% of offspring; the authors also note that superovulation, artificial insemination and cloning allow a popular carrier sire to disseminate an allele to hundreds of animals within a few years, and that AQHA rules from 2007 excluded homozygous HYPP (H/H) horses from registration (Direct; PMID: 19119976). The HYPP disease illustrates how a founder sire plus selection disseminated a dominant disease allele through the breed (Direct; PMID: 1338908). HERDA adds a caution that pedigree-based risk can be underestimated: Bayesian threshold analysis of 1,295 Quarter Horses with 58 affected horses gave a heritability of 0.38 +/- 0.13, supported a single autosomal recessive mode, and estimated carrier frequencies of 1.84% among mares bred to carrier stallions overall but 6.45% within one popular stallion's line; affected horses also had roughly four-fold higher mean inbreeding coefficients than controls (Direct; PMID: 15822588). Homozygosity mapping subsequently identified the PPIB c.115G>A missense variant as the causal candidate for HERDA (Direct; PMID: 17498917). GBED, a fatal glycogen storage disease of Quarter Horse foals (Direct; PMID: 17063718), is caused by a nonsense GBE1 mutation; a random-sample survey found mutant GBE1 allele frequencies of 0.041 in registered Quarter Horses (8.3% heterozygotes), 0.036 in Paints (7.1%) and 0.000 in Thoroughbreds, with homozygotes representing approximately 2.5% of fetal and early neonatal deaths of Quarter Horse lineage at two diagnostic laboratories, most as abortions, and pedigree data supporting inheritance from a single founder stallion (Direct; PMID: 17063718). The PSSM1 GYS1 mutation was identified in PSSM-diagnosed horses across 17 breeds, with the highest prevalence among Draft (87%) and Quarter Horse-related (72%) breeds under grade-2 diagnostic criteria, versus 18% in Warmbloods, and no mutation detected among diagnosed Standardbreds, Thoroughbreds or Arabians (Direct; PMID: 18691366). In the Quarter Horse, the E321G missense in MYH1 has a whole-breed allele frequency of 0.034 (6.8% of individuals) with subgroup maxima of 0.135 in reining horses, making it more prevalent than the HYPP or HERDA alleles in the general breed (Direct; PMID: 30623495). In Warmbloods, screening of 4,081 horses across 38 breeds found 4.9% WFFS (PLOD1 c.2032G>A) carriers overall, with carrier frequencies as high as ~17% in Hanoverian and Danish Warmbloods, about 11% averaged across Warmbloods, and 2.37% in Thoroughbreds, while the allele was absent in most non-Warmblood breeds (Direct; PMID: 33353040). Country-level estimates include an Irish sport horse carrier frequency of 1.98% and Thoroughbred 2.75% (Direct; PMID: 34663462), a Hanoverian allele frequency of 0.14 tracing most carriers to a single F/W-line stallion born in 1861 (Direct; PMID: 32323341), and confirmation that the allele is present at only ~1.2% allele frequency (2.4% carriers) in US Thoroughbreds and is not associated with catastrophic breakdown (Direct; PMID: 31502696). Critically, characterization of abortions, stillbirths and non-viable foals showed that the predominant manifestation is full-term live birth of non-viable foals rather than early abortion: in 14 homozygous WFFS foals, most were born alive, never attempted to stand, and 12 had visible skin defects (Direct; PMID: 31785623). Other well-validated lethal recessives with carrier data include malignant hyperthermia (RYR1 c.7360C>G), detected exclusively in Quarter Horse-related breeds at very low frequency (0.0013 in Quarter Horses, 0.0012 in Paints and Appaloosas, no homozygotes among 159,227 tested horses) but fatal in 51% of phenotyped heterozygotes after anaesthetic or non-anaesthetic triggers (Direct; PMID: 40298507); Arabian-specific cerebellar abiotrophy, severe combined immunodeficiency and lavender foal syndrome, for which a MENA-region survey found carriers for CA and SCID but none for LFS (Direct; PMID: 34946842), while the LFS-causing MYO5A exon 30 deletion was found in 10.3% of sampled Egyptian Arabians (Direct; PMID: 20419149); overo lethal white syndrome caused by the EDNRB mutation in Paint Horses (Direct; PMID: 9585428), for which high-resolution melting assays have been validated for carrier detection (Direct; PMID: 30096357); and foal immunodeficiency syndrome in Fell and Dales ponies (Direct; PMID: 21750681). Across these disorders, the consistent breeding message in the literature is that whole-breed frequencies mask order-of-magnitude subgroup differences, so mate selection should be guided by test results for the specific sire lines involved, not by breed averages (Direct; PMID: 19119976).

Coat colour and pattern genetics: the test panel, prediction accuracy, and its limits

Base coat colour is governed by a small, well-validated panel. A missense mutation in MC1R is associated with chestnut (Direct; PMID: 28721125), and MC1R and ASIP interact synergistically to produce black, brown, dark bay, bay and chestnut, with the EEEE genotype frequency at MC1R decreasing from dark to light colours and the AAAA ASIP genotype frequency increasing as coats lighten (Direct; PMID: 30785190). Dilution phenotypes map to additional loci: SLC45A2 alleles explain cream and, in compound heterozygotes, the pearl and novel sunshine dilutes, with complete genotype-phenotype concordance in testing (Direct; PMID: 31006892); SLC36A1 exon 2 carries the champagne variant (Direct; PMID: 18802473); and the dun dilution maps to TBX3, where Konik data show incomplete penetrance and an additional allele influencing the degree of dilution (Direct; PMID: 34432873). White spotting and patterning loci are more complex: leopard complex spotting and congenital stationary night blindness are caused by a 1,378 bp endogenous retroviral LTR insertion in intron 1 of TRPM1, completely associated with both phenotypes and shown to disrupt transcription by premature polyadenylation (Direct; PMID: 24167615), consistent with earlier fine-mapping of TRPM1 (Direct; PMID: 20353955); a large deletion spanning exons 6-9 of MITF (~63 kb) causes a splashed-white-like phenotype with an increased risk of deafness in an American Paint Horse family (Direct; PMID: 30644113); curly coat arises from variants in both KRT25 and SP6, with mutant KRT25 masking the SP6 allele effect (epistasis) and KRT25 carriers additionally showing hypotrichosis (Direct; PMID: 29686323); and the economically important grey phenotype maps to equine chromosome 25 (Direct; PMID: 12354140). Simultaneous multi-locus genotyping of these colour genes has been developed for registration use (Direct; PMID: 19912416). Prediction is accurate for simple base colours but demonstrably imperfect for patterned phenotypes: in 1,195 American Paint Horses digitally phenotyped for percent white coat, white percentage increased additively with the number of spotting alleles carried regardless of locus, horses with a chestnut (e/e) base had significantly higher white percentages, and even the best multi-locus prediction model explained at best about 55% of phenotypic variation (Direct; PMID: 38002953). Visual recording itself is a major error source: in the semi-feral Sarcidano population, genetic versus registered coat colour showed a 53.4% error rate, with many visually recorded Bays actually genetically Black (Direct; PMID: 38396512), and TBX3/ASIP genotyping of Hucul horses revealed discrepancies between predicted and officially recorded colours in roughly a third of individuals, implicating incomplete penetrance and modifier loci (Direct; PMID: 30614426). A genome-scale survey of 605 horses adds further caution about single-locus interpretation: the SLC45A2 sunshine variant, reported as a rare dilute, is the major allele (>0.86 frequency) in Arabians, Belgians, Morgans, Thoroughbreds and Quarter Horses, and several KIT white-marking and roan variants are common across many breeds at frequencies inconsistent with their originally reported phenotypes (Direct; PMID: 38600096).

Performance genetics: myostatin, race distance, and how far the prediction generalises

The best-established performance marker is the MSTN locus. In Thoroughbreds, the intron 1 SNP g.66493737C>T is strongly associated with best race distance (P = 4.85 x 10^-8 in elite race winners), with the C allele roughly twice as frequent in short-distance as in long-distance cohorts (0.72 vs 0.36), mean best race distances of 6.2, 9.1 and 10.5 furlongs for C/C, C/T and T/T horses respectively, and two-year-old C/C and C/T animals earning substantially more than T/T horses (Direct; PMID: 20098749). A genome-wide SNP association study independently confirmed the ECA18 peak, with BIEC2-417495 the most significant SNP (P = 1.61 x 10^-9) and the highest linkage disequilibrium with the MSTN SNP (r^2 = 0.86) (Direct; PMID: 20932346). Mechanistic work explains why the marker works: the 227 bp SINE insertion in the MSTN promoter is in full concordance with the C allele in Thoroughbreds, and reporter assays plus transcription start site mapping show the SINE alone shifts the TSS into the insertion and reduces myostatin production about 4.5-fold, while the SNP itself has no measurable effect (Direct; PMID: 30379863); the insertion also generates a novel CpG island that could mediate methylation-based regulation (Direct; PMID: 26100061); and in vivo, serum myostatin is genotype-dependent, with homozygous SINE horses at markedly lower median concentrations than heterozygotes and wild-type horses (Direct; PMID: 33846367). MSTN genotype additionally associates with skeletal muscle mRNA levels and post-training expression changes (Direct; PMID: 22497477) and with measured speed indices in horses in training (Direct; PMID: 22016373). However, the same literature shows the prediction is breed- and context-dependent rather than deterministic. The SINE insertion is common only in Thoroughbreds (0.48) and Quarter Horses (0.73) and rare elsewhere, while the intron 1 C allele occurs in many breeds without the SINE; critically, in Belgian draft horses, which lack the SINE, the intron 1 SNP shows no association with muscle fibre type proportions, and the breed effect on fibre type disappears when SINE-negative horses alone are compared - evidence that the SINE, not the SNP, is the functional allele under selection (Direct; PMID: 25160752). In Anglo-Arabians, where the SINE is not detected, the MSTN SNP is nonetheless polymorphic and heterozygous (TC) horses outperformed TT homozygotes on several race-performance indices, an association the authors tentatively attribute to heterosis (Direct; PMID: 33808485). In a Japanese cohort of 1,710 Thoroughbreds, ECA18 SNPs influenced performance rank and win-race distance in both sexes but with clear sex-specific patterns, and the association with earnings reflected the distance distribution of Japanese racing (Direct; PMID: 22221024). MSTN also has pleiotropic effects beyond racing: conformation and gait EBVs in Icelandic horses associate with MSTN-region SNPs (Direct; PMID: 27208149). Beyond MSTN, two-stage GWAS of racing performance in 1,156 Thoroughbreds identified 28 associated SNPs implicating 17 genes, many involved in myogenesis and muscle maintenance (Direct; PMID: 25925054), a harness-racing GWAS in the Norwegian-Swedish Coldblooded Trotter pointed instead to genes for learning and energy metabolism (Direct; PMID: 30105047), and Quarter Horse subpopulations show breed-specific selection signatures on performance genes (Direct; PMID: 30105047). Finally, the three Thoroughbred bone-fracture risk variants at/near MSTN are themselves the major alleles (51.3-55.3%) in Thoroughbreds and present in all breed groups in a 605-horse catalogue, implying that if they contribute to fracture risk their effect sizes are small - a caution against over-interpreting any single marker (Direct; PMID: 38600096).

Measuring genetic diversity and inbreeding: methods and reported levels across breeds

Modern equine diversity work rests on genome-wide SNP platforms, beginning with the 54,602-marker EquineSNP50 BeadChip, which documented breed differentiation, the highest linkage disequilibrium and inbreeding coefficients in the Thoroughbred (F ~ 0.15), and successful mapping of coat-colour loci as a proof of principle (Direct; PMID: 22253606). The field now routinely compares three classes of metric: pedigree-based inbreeding (Fped), genomic excess-homozygosity measures, and runs-of-homozygosity-derived coefficients (FROH) subdivided by ROH length to date inbreeding events. Reported levels vary widely by breed history. The endangered Sorraia shows low diversity and inbreeding of conservation concern (Direct; PMID: 17404326). In three Sicilian populations, genomic inbreeding was highest in Arab (FROH 0.19) and Purosangue Orientale Siciliano (0.13) and lowest in the Siciliano (0.07), while contemporary effective population sizes were critically low for the Purosangue Orientale Siciliano (cNe 10) and Sanfratellano (31); ROH length classes placed much Siciliano autozygosity in the distant past and a substantial fraction of Arab and Sanfratellano inbreeding within roughly the last 70 years (Direct; PMID: 35445758). The isolated Sable Island feral population shows the highest FROH in a multi-breed comparison (mean 0.29, up to 0.42) with a very high ROH count, yet only ~4% of its genome in ROH >16 Mb, indicating inbreeding driven mainly by historical bottlenecks and founder effects rather than recent close-relative mating (Direct; PMID: 35820826). The Rhenish German Draught Horse illustrates why pedigree and genomic estimates diverge: FROH averaged 0.099 versus Fped of only 0.016, with pedigree depth the likely limiting factor; recent effective population size derived from linkage disequilibrium fell to ~59 in the last generation, and ROH islands on seven chromosomes harboured genes for morphology (HOXB cluster), fertility (AURKC, NLRP5, DLX3), muscle growth and skin physiology (ZNF clusters) (Direct; PMID: 40149478). Similar genomic-versus-pedigree discrepancies with moderate genomic inbreeding are reported for the Mangalarga Marchador (Direct; PMID: 32949478) and the Italian Heavy Draught Horse, where pedigree and genomic inbreeding were nevertheless strongly correlated (Direct; PMID: 32751586). At the cross-species level, the median Ne/Nc ratio across livestock breeds is about 0.03, roughly an order of magnitude below wild mammals, and census-based multipliers can provisionally predict Ne for horse breeds (Direct; PMID: 27160794). Functionally, ROH analysis now extends to fitness: in Pura Raza Espanola mares, whole-genome homozygosity was negatively correlated with fertility EBVs, and 13 of 41 ROH islands were associated with fertility, encompassing genes previously tied to oocyte and embryo development (Direct; PMID: 35252415); a public repository of ROH islands across species, including horses, catalogues selection signatures such as the MSTN region (Direct; PMID: 33397285); and five native Italian breeds show ROH islands overlapping QTLs for morphology, hoof health and osteochondrosis (Direct; PMID: 40380299). Genome-scale burden data add nuance to the diversity-inbreeding relationship: median predicted genetic burden per horse is 730 variants (IQR 613-829), breed means range from 607 (Thoroughbred) to 860 (Icelandic), and burden measures correlate significantly but weakly with breed effective population sizes - a correlation the authors note runs opposite to the direction expected, possibly reflecting the Thoroughbred-derived reference genome inflating burden estimates in distant breeds and under-detecting Thoroughbred-specific variants (Direct; PMID: 38600096).

Complex diseases: heritability evidence, candidate loci, and strength of evidence

Metabolic disease. Equine metabolic syndrome shows a clear heritable component: SNP-based heritabilities for EMS-related traits in Welsh ponies and Morgan horses were moderate to high for most of nine traits measured, including insulin, insulin after an oral sugar test, NEFA, triglycerides, leptin and adiponectin, providing the first direct evidence of genetic contribution to these phenotypes (Direct; PMID: 30472742). Candidate loci are less settled. A GWAS in Arabians identified risk markers near FAM174A, including an 11-guanine 3'UTR polymorphism correlated with elevated insulin and body condition (Direct; PMID: 28380523), but a small pony cohort found the 11G risk allele in only a minority of EMS-affected ponies and no significant association with any obesity or insulin-dysregulation measure, leading the authors to conclude it is unlikely to be a singular key polymorphism for EMS in ponies (Direct; PMID: 32169753). A metabogenomic approach in Arabians - metabolome-wide LC/MS profiling against a composite EMS score - identified four EMS-correlated metabolite features and four candidate regions containing 63 genes, with the strongest signal on ECA19 spanning six genes, no coding variants detected (implying regulatory variation), and top markers capturing 43.52% of EMS score variation; putative metabolites implicated glycerolipid metabolism and purine oxidation (Direct; PMID: 32659253). A locus-level analysis showed the HMGA2 height allele (A allele frequency 0.76 in Welsh ponies; genotype-height correlation -0.75) is also correlated with insulin, post-OST insulin, NEFA and triglycerides, explaining ~40% of height and ~20% of insulin heritability and supporting a pleiotropic height-metabolism architecture in ponies (Direct; PMID: 30666754). A deep multi-omic cohort study identified two metabotypes, one defined by elevated primary bile acids (notably taurocholic acid) associated with higher post-challenge insulin, and observed an inverse relationship between genetic diversity (ROH) and insulin levels, suggesting a genetic component to MetS susceptibility (Direct; PMID: 41310118). A suggestive GWAS in Finnhorses nominated candidate loci for insulin dysregulation including ARGLU1 and EFNB2 (Direct; PMID: 38232813). For laminitis specifically, an observational pony cohort found that a previous laminitis diagnosis was consistent with a dominant major gene with reduced penetrance, and that a 'prelaminitic metabolic syndrome' profile (body condition, triglycerides and insulin proxies) predicted most subsequent clinical cases of pasture-associated laminitis, the first such data-supported risk-factor set in equids (Direct; PMID: 16677122).

Musculoskeletal and gastrointestinal disease. In UK Thoroughbreds, distal limb fracture heritability ranged 0.21-0.37 and superficial digital flexor tendon injury 0.31-0.34 on the logistic scale, with a significant positive genetic correlation between the two conditions (0.44 overall; 0.72 within jump racers, non-significant in flat racers), implying that selection against one condition could reduce risk of the other (Direct; PMID: 24679457); an earlier Bayesian estimate for SDFT injury of ~0.17-0.19 (Direct; PMID: 19134077). A Hong Kong cohort reported smaller heritabilities (roughly 0.01-0.20 depending on trait and model) for fracture, osteoarthritis, suspensory ligament injury and tendon injury, with positive genetic correlations between fracture and both osteoarthritis (0.85) and suspensory ligament injury (0.59), and the authors flag that free-text diagnoses and data autoregression limit precision (Direct; PMID: 23746478). For osteochondrosis, tarsal OC risk loci on equine chromosome 14 were detected in US Standardbreds and validated in an independent Norwegian Standardbred population, although breed specificity remains to be determined (Direct; PMID: 26753841). Degenerative suspensory ligament desmitis in the Peruvian Horse has a pedigree-based heritability of 0.22 +/- 0.08, a polygenic architecture with 151 associated SNPs at a permutation threshold and three Bonferroni-significant loci (NOG, AHR, UBE3D), a TGFB3 selection signature unique to cases, enrichment of proteoglycan metabolism, extracellular matrix homeostasis and hedgehog signalling pathways, and a promising Bayesian ridge regression polygenic risk score (R^2 ~ 0.70 on the liability scale) - though the authors caution that validation in larger independent sets is needed (Direct; PMID: 37645058). Complete-rotation large colon volvulus in Kentucky Thoroughbred broodmares appears moderately heritable from a preliminary sire-model estimate, and the authors stress the wide standard error (Direct; PMID: 31391286). Equine grass sickness showed significant familial clustering and affected-sibling odds elevation in a Hungarian stud, with a tentative suggestion of X-linked recessive inheritance that remains unconfirmed (Direct; PMID: 33128520). Infectious- and immune-modified complex traits also show host genetic influence: sarcoid susceptibility maps to an intronic DQA1 (MHC class II) SNP on ECA20 and to a four-gene region on ECA22 with odds ratios of 5.05 and 4.02 respectively, a pattern paralleling MHC associations with HPV-induced tumours in humans (Direct; PMID: 27037728); insidious uveitis in Appaloosas is highly heritable, with the LP/TRPM1 locus explaining only part of the heritability and a further locus on the X chromosome newly associated (Direct; PMID: 41068571); the DDB2 T338M recessive risk factor for ocular squamous cell carcinoma, established in Haflingers, Belgians and Rocky Mountain Horses, extends to additional breeds including Connemara ponies while being absent in Arabians and Appaloosas (Direct; PMID: 33291392; Direct; PMID: 31637255); equine asthma shows environmental and possible maternal-line genetic effects in Polish Koniks (Direct; PMID: 34438743); and cervicothoracic vertebral malformations aggregate within a Thoroughbred-descended Warmblood family, providing the first familial (though not genetically mapped) evidence for inheritance (Direct; PMID: 37958170).

Gaps, misclassification risk, and synthesis

The strongest evidence gap is quantitative: causal variants exist for fewer than 20% of suspected equine genetic traits, and healthy horses carry a high predicted genetic burden - a median of 730 high-impact variants per horse (IQR 613-829), roughly 1.4-2.6-fold higher than human estimates - so variants found in many unrelated healthy horses are unlikely to be simple Mendelian causes (Direct; PMID: 38600096). The same catalogue documents concrete misclassifications: the five reported congenital liver fibrosis variants occur in over nine breeds at frequencies exceeding 89% in some, with 369/605 horses carrying at least one and five carrying all five; the juvenile idiopathic epilepsy variant is the major allele in Standardbreds and Arabians despite later studies showing no association; and all six commercially marketed PSSM2/myofibrillar myopathy variants are present in multiple breeds at frequencies above 5%, making simple Mendelian contribution unlikely (Direct; PMID: 38600096). Independent evaluation confirms the practical consequence: commercial PSSM2/MFM tests (P2, P3a/b, P4) showed no significant association with a histopathological diagnosis in Warmbloods or Arabians, and the variants are also present in ancient domestic horses (Direct; PMID: 32896939). Similar failures of candidate associations to replicate include the FAM174A 11G EMS locus in ponies (Direct; PMID: 32169753) and the KCNE4 anhidrosis variant, which failed validation in a well-phenotyped herd despite a high allele frequency overall (Direct; PMID: 39953936). Methodological caveats recur throughout: the equine reference genome derives from a Thoroughbred, which can inflate apparent burden in distant breeds and hide Thoroughbred-specific variants, motivating a pangenome (Direct; PMID: 38600096); pedigree-based inbreeding routinely underestimates genomic inbreeding where pedigree depth is limited (Direct; PMID: 40149478); small within-breed sample sizes limit power and ascertainment (Direct; PMID: 38600096; Direct; PMID: 35445758); and several complex-disease heritability estimates carry wide standard errors and phenotypic misclassification risk (Direct; PMID: 23746478; Direct; PMID: 31391286).

Synthesis and outlook. The field is strongest where a single, validated, genotype-phenotype-concordant variant meets a clearly stratified breed population - the Quarter Horse disease alleles, WFFS in Warmbloods, the Arabian lethal trio, TRPM1/LP - and here the literature consistently supports test-guided mate selection over breed-average risk, because subgroup carrier frequencies can exceed whole-breed averages by an order of magnitude (Direct; PMID: 33353040). Coat colour prediction is reliable for base colours but degrades for multi-locus white patterning, dilution combinations and visually ambiguous phenotypes, where registries themselves introduce error (Direct; PMID: 38396512; Direct; PMID: 38002953). Performance prediction has one mechanistically grounded, highly predictive locus (MSTN) whose functional allele and limits of cross-breed transfer are now well characterised (Direct; PMID: 30379863; Direct; PMID: 25160752), with the remaining heritable performance variance polygenic and breed-specific (Direct; PMID: 25925054). Diversity monitoring has converged on ROH-based methods that reveal history invisible to pedigrees and connect homozygosity to fertility (Direct; PMID: 35252415; Direct; PMID: 40149478). For complex diseases, heritability is demonstrable for metabolic traits, fracture, tendon injury, DSLD and ocular conditions, but causal variant discovery has largely stalled at candidate regions, risk loci that fail replication, and preliminary estimates; the most promising current directions are within-breed polygenic risk modelling (Direct; PMID: 37645058), metabolomic endophenotyping integrated with genomics (Direct; PMID: 32659253; Direct; PMID: 41310118), and systematic population-level re-evaluation of marketed variants against the genomic burden baseline before new tests are commercialised (Direct; PMID: 38600096; Direct; PMID: 32896939).

Research notebook

Which inherited diseases do horse breeders test for (HERDA, HYPP, GBED, etc.), what mutations are involved, what carrier frequencies exist, and how should carrier status guide mating choices?

Quantitative carrier frequencies from elite-subgroup genotyping of American Quarter Horses show extreme stratification: overall allele frequencies were PSSM1 0.055, GBED 0.054, HERDA 0.021, HYPP 0.008 (Paint Horse OLWS/EDNRB 0.107), but within subgroups HYPP reached 0.299 in halter horses (56.4% carriers), PSSM1 0.155 in halter horses (28.2% carriers), HERDA 0.142 in cutting horses (28.3% carriers) and GBED 0.132 in western pleasure horses — so whole-breed frequencies grossly underestimate risk for specific performance lines, and recessive carrier-to-carrier matings produce affected foals 25% of the time while dominant heterozygotes pass the allele to 50%.

Status: verified • Confidence: high

• Hyperkalaemic periodic paralysis in Quarter Horses was shown to be linked to the horse adult skeletal muscle sodium channel gene, identifying SCN4A as the genetic basis of the disease. (PMID 1323940, abstract)
• The periodic paralysis mutation in Quarter Horses is a sodium channel mutation disseminated by selective breeding, documenting that a founder sire and breeding choices spread the disease allele through the breed. (PMID 1338908, abstract)
• Homozygosity mapping identified a missense mutation in equine cyclophilin B (PPIB) associated with HERDA in the American Quarter Horse, establishing the causal variant and an autosomal recessive mode of inheritance. (PMID 17498917, abstract)
• The inheritance of hereditary equine regional dermal asthenia in Quarter Horses was analysed and found consistent with an autosomal recessive trait, defining the mating risk that heterozygote pairing produces affected foals. (PMID 15822588, abstract)
• Evaluation of allele frequencies of inherited disease genes in subgroups of American Quarter Horses reported that HERDA carrier frequency is elevated in cutting-line subgroups compared with other Quarter Horse subpopulations, alongside frequencies for HYPP, GBED and PSSM. (PMID 19119976, abstract)
• The HERDA-causing PPIB mutation was detected among Quarter Horses in France, demonstrating the mutation has spread to European Quarter Horse populations. (PMID 21118319, abstract)
• Allele frequency of HERDA in American Quarter Horses in Brazil was quantified by high-resolution melting PCR, confirming the variant occurs outside North America and enabling carrier screening there. (PMID 24355815, abstract)
• A glycogen branching enzyme (GBE1) mutation was shown to cause equine glycogen storage disease type IV, identifying GBE1 as the causal gene for GBED. (PMID 15366377, abstract)
• Glycogen branching enzyme deficiency was described in Quarter Horse foals as a fatal disease presenting with weakness, respiratory failure and sudden death, characterising the clinical phenotype of homozygotes. (PMID 11817063, abstract)
• The allele frequency of the GBED-causing GBE1 mutation in Quarter Horse and Paint Horse populations was estimated and its likely population impact assessed, indicating the deleterious allele is maintained at a measurable frequency in these breeds despite embryonic/foal lethality of homozygotes. (PMID 17063718, abstract)
• A glycogen synthase 1 (GYS1) mutation was found to cause a novel skeletal muscle glycogenosis (type 2 polysaccharide storage myopathy) and was identified in diverse breeds, extending PSSM1 testing beyond Quarter Horse-related breeds. (PMID 18691366, abstract)
• A missense mutation in MYH1 (E321G) was shown to be associated with susceptibility to immune-mediated myositis in Quarter Horses, providing a DNA test target for a breed-associated muscle disease. (PMID 29510741, abstract)
• The prevalence of the E321G MYH1 variant was estimated in performance subgroups of American Quarter Horses, quantifying how common the immune-mediated myositis risk allele is within the breed. (PMID 30623495, abstract)
• Analysis of a neuromuscular disease laboratory submission population reported the prevalence of known genetic mutations (including PSSM1, HERDA, GBED and MYH1 variants) in horses tested for muscle disease, reflecting real-world use of these tests in diagnostics and breeding screening. (PMID 36150530, abstract)
• A neonatal foal with severe skin malformations tested homozygous positive for the Warmblood Fragile Foal Syndrome variant, confirming the genotype-phenotype match for the PLOD1 mutation. (PMID 25637337, abstract)
• Distribution analysis of the WFFS type 1 mutation (PLOD1 c.2032G>A) across different horse breeds from Europe and the United States found the variant occurs at low frequency in multiple Warmblood breeds while being absent or very rare in non-Warmblood breeds. (PMID 33353040, abstract)
• The WFFS PLOD1 variant was found to have a low allele frequency in the Thoroughbred breed and was not associated with catastrophic breakdown injuries, indicating the mutation is largely confined to Warmblood populations. (PMID 31502696, abstract)
• Estimation of the WFFS causative SNP frequency in horses in Ireland provided a national carrier-frequency figure used to inform breeding recommendations. (PMID 34663462, abstract)
• The Hanoverian F/W mare line was shown to contribute to segregation of the WFFS variant (PLOD1 c.2032G>A) in Warmblood horses, tracing the mutation to a specific maternal family. (PMID 32323341, abstract)
• Foals, abortions and stillbirths homozygous for the WFFS variant were characterised, documenting the full range of reproductive failure outcomes of carrier-to-carrier matings beyond liveborn affected foals. (PMID 31785623, abstract)
• A missense mutation in the endothelin-B receptor gene was associated with Lethal White Foal Syndrome, an equine version of Hirschsprung disease, establishing the EDNRB test for Paint Horse breeders. (PMID 9585428, abstract)
• High-resolution melting analysis was validated as a diagnostic tool for the endothelin receptor B (EDNRB) mutation in American Paint Horses and used to estimate allele frequency, supporting routine carrier screening. (PMID 30096357, abstract)
• A mutation associated with fatal foal immunodeficiency syndrome was identified in the Fell and Dales ponies, adding a breed-specific lethal recessive test. (PMID 21750681, abstract)
• Cerebellar abiotrophy, Lavender Foal Syndrome and severe combined immunodeficiency variants were surveyed in three MENA-region breeds, providing carrier-frequency data for Arabian-related populations. (PMID 34946842, abstract)
• A whole-genome SNP association approach identified a deletion in myosin Va (MYO5A) responsible for Lavender Foal Syndrome, a lethal coat-colour-linked disease of Arabian foals. (PMID 20419149, abstract)
• Allele frequencies and genotypes for the ryanodine receptor 1 (RYR1) variant causing malignant hyperthermia and fatal rhabdomyolysis with hyperthermia in horses were reported, informing breeding advice for this anaesthesia-triggered disease. (PMID 40298507, abstract)
• Full-text genotyping of elite Quarter Horse subgroups reported whole-breed allele frequencies of PSSM1 0.055, GBED 0.054, HERDA 0.021 and HYPP 0.008 (Paint Horse LWFS 0.107), but subgroup frequencies of HYPP 0.299 in halter horses (56.4% carrying), PSSM1 0.155 in halter horses (28.2% carrying), HERDA 0.142 in cutting horses (28.3% carrying) and GBED 0.132 in western pleasure horses, and estimated 90,720-181,440 AQHs carrying HERDA, 278,640-421,200 carrying GBED and 272,650-419,600 carrying PSSM1 breed-wide. (PMID 19119976, results)
• The study's discussion explains that recessive carrier-to-carrier matings produce affected foals 25% of the time, dominant heterozygotes pass the allele to 50% of foals, AQHA rules from 2007 excluded homozygous HYPP horses from registration, and that with artificial insemination and cloning a few years suffice to spread a disease allele from a popular unscreened sire to thousands of offspring — hence thorough screening of popular breeding stock is recommended. (PMID 19119976, discussion)
• Allele frequency of HERDA in American Quarter Horses in Brazil was quantified by high-resolution melting PCR, providing a non-US carrier frequency estimate. (PMID 24355815, abstract)
• Allele frequencies and genotypes for the RYR1 malignant hyperthermia variant were reported, extending frequency data to another testable disease allele. (PMID 40298507, abstract)

Which genes underlie horse coat colours and patterns, how accurate are commercial genetic tests at predicting foal colour, and what are the known limitations (unaccounted modifiers, epistasis, misrecorded phenotypes)?

The grey coat colour, one of the most economically important colour traits in breeds like the Thoroughbred, maps to the grey locus on horse chromosome 25 (later resolved as a duplication in STX17 region), and its frequency differs between racing subpopulations — completing the core panel of base-colour, dilution and white-pattern loci that commercial tests cover.

Status: verified • Confidence: medium

• A missense mutation in the melanocyte-stimulating hormone receptor gene (MC1R) was found to be associated with the chestnut coat colour in horses, defining the extension locus. (PMID 8995760, abstract)
• Synergy between MC1R and ASIP was demonstrated for coat colour in horses, showing these two loci interact epistatically to produce bay, black and chestnut phenotypes. (PMID 30785190, abstract)
• A method for simultaneously genotyping multiple horse coat colour loci was developed and applied to basic colour variation in Thoroughbreds and Misaki horses, enabling multi-locus colour prediction. (PMID 19912416, abstract)
• Candidate gene analysis identified variants in SLC45A2 that explain the pearl and sunshine dilute phenotypes in compound heterozygote horses, refining cream-locus genotype-to-phenotype prediction. (PMID 31006892, abstract)
• A missense mutation in exon 4 of SLC45A2 was confirmed as the cause of the pearl coat dilution in horses. (PMID 30968968, abstract)
• A missense mutation in exon 2 of SLC36A1 was shown to be responsible for champagne dilution in horses, adding another dilution locus to the test panel. (PMID 18802473, abstract)
• Fine-mapping and mutation analysis of TRPM1 established it as the candidate gene for leopard complex spotting and congenital stationary night blindness in horses. (PMID 20353955, abstract)
• A retroviral insertion in TRPM1 was shown to be the cause of congenital stationary night blindness and leopard complex spotting in the horse, defining the precise causal variant. (PMID 24167615, abstract)
• Whole-genome sequencing revealed a large deletion in the MITF gene in horses with white spotted coat colour and increased risk of deafness, linking a colour test result to a health outcome. (PMID 30644113, abstract)
• Analysis of coat colour variation in the Konik horse provided new insight into the phenotypic effect of the dun dilution variant (dun versus non-dun). (PMID 34432873, abstract)
• Digital phenotyping of American Paint Horses revealed phenotype diversity and epistasis among white spotting alleles, showing that the visible pattern depends on interactions between spotting loci rather than single-gene effects. (PMID 38002953, abstract)
• TBX3 and ASIP genotypes revealed discrepancies in officially recorded coat colours of Hucul horses, documenting that registry colour records can disagree with genotype-based prediction. (PMID 30614426, abstract)
• Discrepancies between genetic and visual coat colour assignment were documented in the Sarcidano horse, an example of limits to genotype-based colour prediction. (PMID 38396512, abstract)
• An epistatic effect of KRT25 on SP6 was shown to be involved in curly coat in horses, illustrating that coat-type traits require multi-locus interpretation. (PMID 29686323, abstract)
• Full-text re-analysis of 605 genomes found the SLC45A2 'sunshine' variant is the major allele (>0.86 frequency) in Arabians, Belgians, Morgans, Thoroughbreds and Quarter Horses despite being reported to cause a rare dilute phenotype, and found KIT white-marking (W20, W31, W34) and roan variants distributed across many breeds, with homozygotes for several roan-associated variants — concluding further phenotype investigation is warranted for these marketed colour variants. (PMID 38600096, discussion)
• Digital phenotyping of American Paint Horses revealed phenotype diversity and epistasis among white spotting alleles, showing the visible pattern depends on interactions between spotting loci. (PMID 38002953, abstract)
• TBX3 and ASIP genotypes revealed discrepancies in officially recorded coat colours of Hucul horses, documenting registry colour records that disagree with genotype-based prediction. (PMID 30614426, abstract)
• Discrepancies between genetic and visual coat colour assignment were documented in the Sarcidano horse. (PMID 38396512, abstract)
• An epistatic effect of KRT25 on SP6 was shown to be involved in curly coat in horses, illustrating multi-locus coat traits. (PMID 29686323, abstract)
• The grey coat colour locus was linked to microsatellites on horse chromosome 25, positioning the grey gene for subsequent fine-mapping. (PMID 12354140, abstract)
• The roles of the Gray and Extension loci were evaluated in the coat coloration of Thoroughbred racing horses, quantifying how genotype predicts the grey phenotype in a major breed. (PMID 28721125, abstract)

Which genes and genomic markers are linked to performance traits in horses, and how reliable/predictive are these associations across breeds and cohorts?

The reliability of MSTN-based distance prediction is context-dependent: its effect varies across geographic racing regions and additional modifying loci contribute to distance aptitude; independent cohorts (Japanese Thoroughbreds, Anglo-Arabians, Icelandic horses) show breed-specific associations and pleiotropic effects; genome-scale data show even MSTN-linked fracture-risk variants are major alleles (~51–55%) in Thoroughbreds and present across all breed groups, implying small effect sizes; and multi-locus GWAS identify further performance genes — so commercial 'speed gene' tests are informative but not deterministic and do not transfer unmodified across populations.

Status: verified • Confidence: medium

• A sequence polymorphism in MSTN was shown to predict sprinting ability and racing stamina in Thoroughbred horses, the original description of the 'speed gene'. (PMID 20098749, abstract)
• The myostatin locus (MSTN) was identified as having a major effect on optimum racing distance in Thoroughbred horses in the USA. (PMID 21070290, abstract)
• A genome-wide SNP association study confirmed the g.66493737C>T variant in equine MSTN as the most powerful predictor of optimum racing distance for Thoroughbred racehorses. (PMID 20932346, abstract)
• MSTN genotypes in Thoroughbred horses were shown to influence skeletal muscle gene expression and racetrack performance, connecting the variant to functional muscle biology. (PMID 22497477, abstract)
• MSTN genotype (g.66493737C/T) was associated with speed indices in Thoroughbred racehorses, extending the association to measured racing speed. (PMID 22016373, abstract)
• The 'speed gene' effect of myostatin in Thoroughbred horses was shown to arise from a promoter-proximal SINE insertion, identifying the likely functional variant underlying the race-distance association. (PMID 30379863, abstract)
• A putative regulatory mechanism for the MSTN gene was proposed in which the SINE marker Ins227bp generates a CpG island that regulates MSTN expression. (PMID 26100061, abstract)
• A highly prevalent SINE mutation in the myostatin gene promoter was associated with low circulating myostatin concentration in Thoroughbred racehorses, supporting a functional effect of the insertion. (PMID 33846367, abstract)
• Haplotype diversity in the equine myostatin gene was characterised with focus on variants associated with race distance propensity and muscle fibre type proportions, mapping the haplotype structure across breeds. (PMID 25160752, abstract)
• The contribution of MSTN and additional modifying genetic loci to race distance aptitude was quantified in Thoroughbreds racing in different geographic regions, showing the speed gene explains part but not all of the trait and its effect differs between regions. (PMID 30604488, abstract)
• A cohort study of racing performance in Japanese Thoroughbred racehorses using genome information on ECA18 (containing MSTN) tested the association in an independent national cohort. (PMID 22221024, abstract)
• Polymorphisms at the myostatin gene were analysed for association with sport performance in Anglo-Arabian racehorses, testing MSTN transferability beyond pure Thoroughbreds. (PMID 33808485, abstract)
• Conformation traits and gaits in the Icelandic horse were associated with genetic variants in MSTN, demonstrating pleiotropic effects of the locus in a gaited breed. (PMID 27208149, abstract)
• A joint analysis of a two-stage genome-wide association study for racing performance of 1,156 Thoroughbreds identified multiple muscle-related genes beyond MSTN, indicating a polygenic architecture for racing performance. (PMID 25925054, abstract)
• A genome-wide association study for harness racing success in the Norwegian-Swedish coldblooded trotter revealed genes for learning and energy metabolism, showing different loci underlie performance in trotting breeds. (PMID 30157760, abstract)
• Genome-wide signatures of selection in American Quarter Horse subpopulations revealed genes associated with performance, showing breed-specific selection has acted on different performance-related loci. (PMID 30105047, abstract)
• Full-text genome-scale data showed the three Thoroughbred bone-fracture risk variants at/near MSTN are the major alleles (51.3–55.3% frequency) in Thoroughbreds and are present in all breed groups, indicating that if they contribute to fracture risk their effect sizes are small — a caution for performance/injury marker interpretation. (PMID 38600096, discussion)
• A joint analysis of a two-stage genome-wide association study for racing performance of 1,156 Thoroughbreds identified multiple muscle-related genes beyond MSTN. (PMID 25925054, abstract)

What methods are used to measure genetic diversity and inbreeding in horse breeds (pedigree-based vs genomic, ROH, Fis, effective population size), and what levels have been reported across breeds?

Reported genomic diversity and inbreeding levels vary widely across breeds: endangered and closed populations such as the Sorraia and feral Sable Island horses show very high ROH-based inbreeding and low diversity, draught breeds (Rhenish German Draught) show elevated homozygosity, managed sport breeds (Mangalarga Marchador, Italian breeds) show moderate inbreeding, and ROH islands overlap fertility and selection-related loci — illustrating that breed history and closed studbooks drive measurable diversity differences.

Status: verified • Confidence: medium

• A high-density SNP array for the domestic horse and extant perissodactyla was developed with demonstrated utility for association mapping, genetic diversity and phylogeny studies, providing the genomic tool underlying most modern equine diversity work. (PMID 22253606, abstract)
• Fine-scale estimation of inbreeding rates, runs of homozygosity and genome-wide heterozygosity was performed in the Mangalarga Marchador horse, comparing pedigree-based and genomic measures within a breed. (PMID 32949478, abstract)
• A genome-wide survey of three local horse populations focused on runs of homozygosity patterns, using ROH to quantify genomic inbreeding and distinguish it from pedigree estimates. (PMID 35445758, abstract)
• Effective population sizes were estimated from census and herdbook data in cattle, sheep, horses, pigs and goats, providing pedigree-based Ne benchmarks for horse breeds. (PMID 27160794, abstract)
• Genetic variability in the Italian Heavy Draught Horse was assessed from both pedigree data and genomic information, a direct comparison of the two method classes. (PMID 32751586, abstract)
• Analysis of inbreeding and genetic structure in the endangered Sorraia horse breed documented high inbreeding and low diversity, with direct implications for conservation and management. (PMID 17404326, abstract)
• Runs of homozygosity in Sable Island feral horses revealed the genomic consequences of inbreeding and divergence from domestic breeds, showing extreme homozygosity in an isolated population. (PMID 35820826, abstract)
• Genomic patterns of homozygosity and genetic diversity were quantified in the Rhenish German Draught Horse, reporting ROH-based inbreeding levels in a draught breed. (PMID 40149478, abstract)
• Fine-scale analysis of runs of homozygosity islands affecting fertility in mares identified genomic regions where homozygosity is associated with reduced fertility, linking inbreeding to reproductive outcomes. (PMID 35252415, abstract)
• A publicly available repository of ROH islands revealed signatures of selection in different livestock and pet species, including horses, enabling cross-breed comparison of homozygosity islands. (PMID 33397285, abstract)
• Selection signatures and inbreeding were jointly explored in five native horse breeds, relating ROH patterns to breed history and management. (PMID 40380299, abstract)

What are the main gaps and limitations in equine genetic testing for complex diseases — validation, replication, and missing evidence?

The main gaps and limitations in equine genetic testing are now well documented from full-text genome-scale evidence: causal variants exist for less than 20% of suspected equine genetic traits; the predicted genetic burden per horse (median 730 high-impact variants, ~1.4–2.6-fold higher than humans) means many 'disease-associated' variants found in healthy horses across many breeds cannot be simple causes — e.g., congenital liver fibrosis candidate variants occur in >9 breeds at frequencies >89% and juvenile idiopathic epilepsy variants are major alleles in unrelated breeds; commercial PSSM2/myofibrillar myopathy tests (COL6A3 K1, MYOT P2, FLNC P3, MYOZ3 P4, PYROXD1 P8) do not correspond to histopathological diagnosis and their variants are common in multiple breeds; and misclassified variants risk horses being excluded from breeding or euthanised.

Status: verified • Confidence: medium

• Heritability of metabolic traits associated with equine metabolic syndrome was estimated in Welsh ponies and Morgan horses, showing measurable heritable variation in EMS-related traits. (PMID 30472742, abstract)
• A genome-wide association study revealed a risk locus for equine metabolic syndrome in the Arabian horse, providing a breed-specific candidate region. (PMID 28380523, abstract)
• A genome-wide association study suggested genetic candidate loci of insulin dysregulation in Finnhorses. (PMID 38232813, abstract)
• Metabogenomics identified four candidate regions involved in the pathophysiology of Equine Metabolic Syndrome, integrating metabolite and genomic data. (PMID 32659253, abstract)
• An HMGA2 variant was evaluated for pleiotropic effects on height and metabolic traits in ponies, linking body-size genetics to metabolic risk. (PMID 30666754, abstract)
• Preliminary analysis of the FAM174A gene suggested it lacks a strong association with equine metabolic syndrome in ponies, an example of candidate loci that failed to replicate. (PMID 32169753, abstract)
• Evaluation of genetic and metabolic predispositions and nutritional risk factors for pasture-associated laminitis in ponies demonstrated that pony laminitis clusters with familial and metabolic predisposition. (PMID 16677122, abstract)
• Bile acids were shown to segregate metabolic syndrome in a cohort of 100 deeply phenotyped horses, adding a metabolomic marker layer to EMS phenotyping. (PMID 41310118, abstract)
• Heritability for superficial digital flexor tendon injury was estimated by Gibbs sampling in the Thoroughbred racehorse, establishing a genetic component to tendon injury risk. (PMID 19134077, abstract)
• Genetic parameters were estimated for distal limb fracture and superficial digital flexor tendon injury in UK Thoroughbred racehorses, quantifying heritable variation in both injury types. (PMID 24679457, abstract)
• Preliminary genetic analyses of important musculoskeletal conditions of Thoroughbred racehorses in Hong Kong estimated heritabilities, with the authors noting the estimates were preliminary. (PMID 23746478, abstract)
• Risk loci for osteochondrosis were identified and validated in Standardbreds by genome-wide association, demonstrating detectable genetic risk factors for a developmental joint disease. (PMID 26753841, abstract)
• The genetic architecture and polygenic risk score prediction of degenerative suspensory ligament desmitis (DSLD) were characterised in the Peruvian Horse, showing a polygenic basis for this breed-associated tendon/ligament disease. (PMID 37645058, abstract)
• An effect of inbreeding on the 'club foot' disorder in Arabian Pureblood horses reared in Italy was reported, linking inbreeding level to expression of a complex conformational disorder. (PMID 31998623, abstract)
• A historic horse family displaying malformations of the cervicothoracic junction was connected to modern German Warmblood horses, illustrating familial aggregation of complex skeletal malformations. (PMID 37958170, abstract)
• Preliminary heritability of complete rotation large colon volvulus was estimated in Thoroughbred broodmares, indicating a possible genetic contribution to a major colic cause. (PMID 31391286, abstract)
• Family aggregation analysis showed a possible heritable background of equine grass sickness (dysautonomia) in a Hungarian stud population, supporting familial clustering without identifying a locus. (PMID 33128520, abstract)
• A DDB2 genetic risk factor for ocular squamous cell carcinoma was identified in three additional horse breeds, extending a known cancer-risk allele across breeds. (PMID 33291392, abstract)
• Additional evidence for DDB2 T338M as a genetic risk factor for ocular squamous cell carcinoma in horses was reported, strengthening the association. (PMID 31637255, abstract)
• A genome-wide investigation of insidious uveitis in Appaloosa horses identified genetic associations, extending complex-disease genomics to immune-mediated ocular disease. (PMID 41068571, abstract)
• Host genetic influence on papillomavirus-induced tumours in the horse was demonstrated, showing genetics modifies susceptibility to infectious tumours. (PMID 27037728, abstract)
• Genetic and environmental effects on equine asthma were analysed in Polish Konik horses, indicating heritable variation in a respiratory complex disease. (PMID 34438743, abstract)
• A genome-wide association analysis of ranked pain responses explored genetic influences on equine analgesic efficacy. (PMID 40216012, abstract)
• Genomic screening detected five new cases of 65,XXY syndrome in the Pura Raza Español breed, showing commercial SNP arrays can also uncover chromosomal abnormalities relevant to fertility. (PMID 39272345, abstract)
• A genome-wide analysis predicted the genetic burden and frequency of phenotype-associated variants in the horse, quantifying how many known deleterious variants each breed carries. (PMID 38600096, abstract)
• Commercial genetic testing for type 2 polysaccharide storage myopathy and myofibrillar myopathy was shown not to correspond to a histopathological diagnosis, a direct example of an unvalidated commercial test. (PMID 32896939, abstract)
• A missense mutation in the KCNE4 gene was shown not to be predictive of equine anhidrosis, another candidate association that failed validation. (PMID 39953936, abstract)
• A survey of health problems in midlife Swiss warmblood horses documented the prevalence of chronic complex conditions, highlighting the disease burden for which genetic tests are lacking. (PMID 33941510, abstract)
• Full-text whole-genome sequencing of 605 horses from 48 breeds quantified a median predicted genetic burden of 730 variants per horse (IQR 613–829), 1.4–2.6-fold higher than humans, with breed means from 607 (Thoroughbred) to 860 (Icelandic) and significant correlations with breed effective population size. (PMID 38600096, results)
• The paper's discussion shows that reported congenital liver fibrosis variants occur across >9 breeds (some >89% frequency), juvenile idiopathic epilepsy variants are major alleles in Standardbreds and Arabians, and all six commercially marketed PSSM2 variants are present in multiple breeds at frequencies >5% — concluding these variants are unlikely to cause these diseases in simple Mendelian fashion and that extensive validation is warranted before tests are marketed, because false positives can lead to mistrust, exclusion from breeding or euthanasia. (PMID 38600096, discussion)
• Commercial genetic testing for type 2 polysaccharide storage myopathy and myofibrillar myopathy was shown not to correspond to a histopathological diagnosis. (PMID 32896939, abstract)
• A missense mutation in the KCNE4 gene was shown not to be predictive of equine anhidrosis, another failed candidate association. (PMID 39953936, abstract)
• The FAM174A gene lacks a strong association with equine metabolic syndrome in ponies, an example of candidate-gene non-replication. (PMID 32169753, abstract)