Cat Breed Genetics

The reference genome used by cat genomics research worldwide comes from an Abyssinian cat named Cinnamon.

The reference genome's role in genomics research is equivalent to the origin point of latitude and longitude on a map. Saying a cat has an SNV at a certain position means that cat differs from Cinnamon at that position. All breed genetic distances, diversity metrics, variant frequencies in disease association studies, all measured from Cinnamon as zero. She was chosen because Abyssinian cats are highly inbred, high homozygosity makes sequence assembly clean. When the cat genome project launched in 2003, sequencing costs meant you could only pick one individual as reference, picking one with high homozygosity reduced the difficulty of handling heterozygous segments during assembly.

The first layer of consequences hits inter-breed comparisons. A study reports that Siamese cats have significantly higher variant frequency at a certain locus than British Shorthairs. How much of that frequency difference is biological difference between the two breeds, and how much is caused by the genetic distance from Siamese to Cinnamon being far greater than from British Shorthair to Cinnamon? Can't tell. No alternative reference genome available for cross-validation. Human genomics solved this with pangenome references, cat genomics doesn't have the resources for a pangenome.

The second layer of consequences hits within-breed analysis. Within-breed genetic diversity estimation depends on polymorphic site identification, polymorphic site identification depends on read alignment to the reference genome. When certain regions of the reference genome diverge too much from the target breed, some reads fail to align or align incorrectly, variants in those regions get underestimated or misestimated. Breeds most genetically distant from Cinnamon are most affected. These highly divergent regions are precisely the ones likely to harbor the greatest functional divergence between breeds.

The third layer hits GWAS. GWAS depends on frequency comparisons between affected individuals and healthy controls across large numbers of SNP loci. If reference genome bias causes systematic overestimation of variants in certain breeds, some association signals in GWAS conducted primarily in those breeds may be artifacts. The total number of cat GWAS is only about a dozen, a systematic assessment of bias might not even be statistically powered.

Microsatellite marker analysis conducted large-scale phylogenetic study on over 1100 cats globally. Persian cat genetic data clusters with European/American group, not Mediterranean. Japanese Bobtail same. Over a century of British and American cattery selective breeding rewrote the genetic chassis. Geographic labels in breed names are cultural labels. Several recently established breeds show no genetic differentiation from their parent breeds at all, from a genomic perspective they are the same thing.

The 2020 publication of the new cat reference genome paper identified a structural variant causing feline dwarfism during structural variation analysis, disrupting the UGDH gene. UGDH encodes UDP-glucose 6-dehydrogenase, catalyzing the oxidation of UDP-glucose to UDP-glucuronic acid. UDP-glucuronic acid is the essential precursor for synthesizing glycosaminoglycans including hyaluronic acid, chondroitin sulfate, and heparin. Glycosaminoglycans are core structural components of cartilage matrix. The mechanism of UGDH loss-of-function causing feline dwarfism is insufficient glycosaminoglycan synthesis leading to defective cartilage matrix assembly, in turn affecting longitudinal bone growth.

This discovery is directly related to the genetic amplification effect of breed inbreeding. In randomly mating populations UGDH deleterious variants exist at extremely low frequency, heterozygous the vast majority of the time, no phenotype. Inbred breeds pushed homozygosity probability up, phenotype exposed. Without breed-driven inbreeding this variant would still be hidden in heterozygous state.

In human genetic disease research large numbers of rare variants are classified as VUS, too few carriers to power statistical tests. Cat breed inbreeding amounts to forcibly homozygosing a rare variant and observing the consequences.

I believe UGDH complete loss-of-function is likely embryonic lethal in humans. The UGDH structural variant in cat breeds causes complete protein truncation, cats survive and are born with dwarfism phenotype. Human gestation is longer, placental development more complex, if dependence on the glycosaminoglycan pathway is greater, complete truncation may not survive the embryonic stage at all. Then the clinically relevant UGDH variants in humans are more likely missense mutations or splice-site variants causing partial loss of function, phenotype not classic dwarfism but milder short stature or cartilage developmental anomalies. These mild phenotypes would not be prioritized for whole-exome sequencing in clinical settings, even if they exist they would be very hard to detect.

Genes encoding other enzymes in the glycosaminoglycan synthesis pathway already have established associations with skeletal dysplasia in humans: CHST3 mutations cause spondyloepiphyseal dysplasia Omani type, CHSY1 mutations cause Temtamy preaxial brachydactyly syndrome. UGDH sits further upstream in this pathway, loss of function theoretically has broader and more severe consequences. The vulnerability of the entire pathway is known, the UGDH node has simply never been checked in humans.

The enrichment of skeletal and cartilage-related genetic anomalies in cat breeds is striking. Scottish Fold osteochondrodysplasia. Munchkin short limbs, long assumed to be similar to canine FGF4 retrotransposon insertion, molecular verification still incomplete. Polydactyl cat Shh gene regulatory region ZRS variant. Each is a high-frequency phenotype maintained within breeds by inbreeding, each corresponds to potentially unidentified disease genes or novel functions of known genes.

The position of the enzyme UGDH encodes in this pathway is worth examining more closely. The reaction UGDH catalyzes is an irreversible four-electron oxidation, the product UDP-glucuronic acid is the common precursor for five different glycosaminoglycan synthesis pathways. This means UGDH functional decline doesn't affect just one glycosaminoglycan, it simultaneously affects synthesis of hyaluronic acid, chondroitin sulfate, dermatan sulfate, heparin, and heparan sulfate. When a downstream pathway-specific enzyme has problems (for instance CHST3 only affects chondroitin sulfate), the phenotype is relatively confined. When UGDH has problems, five pathways are simultaneously affected, both severity and scope of phenotype expand. This may also explain why UGDH complete loss-of-function is more likely lethal in humans while survivable in cats, different species have different dependency weights on these five pathways, feline embryonic development may require lower quantities of certain glycosaminoglycans than human. This is speculation, verifying it would require conditional UGDH knockout in model organisms, nobody has done conditional gene knockout in cats, the technology is not mature.

Dogs went through tens of thousands of years of functional selective breeding, breed differences involve hundreds of genes. Cats have no such history. The cat-human relationship began roughly ten thousand years ago in Fertile Crescent granaries, catching mice, self-domesticated. Breed differentiation didn't start until late 19th century when the British started cat shows, selection targets almost entirely appearance. Every new breed traces back to a single gene mutation in a single cat. Scottish Fold from a white farm cat named Susie in Scotland in 1961, TRPV4-related gene mutation. Selkirk Rex from a shelter cat in Montana in 1987, genetically independent from Cornish Rex, Devon Rex, and German Rex. KIT gene simultaneously controls white coat, blue eyes, and inner ear melanocyte development, congenital deafness in white blue-eyed cats is a consequence of pleiotropic expression. Every breed is a founder effect event, entire population genetic diversity compressed to extreme narrowness.

The mechanism behind breed disease tables is genetic hitchhiking, fixing a target trait simultaneously fixes nearby deleterious recessive alleles. Effective population size Ne, very few cat breeds reach 500. Scottish Fold osteochondrodysplasia is genetically inseparable from the fold ear trait, different in nature from health problems caused by population structure in other breeds, the breed standard itself needs questioning. GCCF has already refused to recognize Scottish Folds.

February 2026 Science published the feline oncogenome study, 493 paired tumor and normal tissue targeted sequencing, 13 cancer types, 31 driver genes. TP53 mutated in 33% of tumors. FBXW7 mutated in over 50% of feline mammary tumors. Papillomavirus sequences detected in some tumors, papillomavirus oncogenic role in cats not yet established. Germline variants are heritable, cancer-predisposing germline variants carried by founder cats become high-frequency within breeds. In dogs, Golden Retrievers have significantly elevated hemangiosarcoma incidence, Bernese Mountain Dog histiocytic sarcoma strongly associated with breed background. The 493 cats in this study came from five countries, unprecedented scale in feline oncology, still small in comparative oncology. Canine comparative oncology has datasets in the thousands. Systematic collection of feline tumor tissue has some infrastructure in the UK and Canada, essentially blank in other countries.