A fantastic genome study of 2000 canids tells a 'story' of selective dog breeding over time and space.
"Incorporating 20 × data from 1987 individuals, including 1611 dogs (321 breeds), 309 village dogs, 63 wolves, and four coyotes, we identify genomic variation across the canid family, setting the stage for detailed studies of domestication, behavior, morphology, disease susceptibility, and genome architecture and function." (Also included revisions of pre-existing data from the earlier Boxer and German Shepard genome sets)
A highlight of this is a "worldwide sampling of village dogs and niche populations, both of which fall outside the umbrella of pure or mixed breed dogs". I'm wondering if the Viking Dogs of Dublin project might have enough genome data of discovered remains to compare with this dataset. It would be fascinating to compare this snapshot from the Midieval dogs.
The wolf data came from 57 wolves from differing geographical areas. I'm a bit disappointed in the limited (4) coyote individuals. Especially given the admixture events in coyotes with dogs and wolves.
"Analysis of mitochondrial data reveals surprisingly few haplotypes in dogs, with greater observed variation in wild canids." What this means is a limited pool of wild canids from which our modern breeds were first selected for domestication.
The analyses included many comparisons: "The major clades [primary groups] are made up of breeds sharing occupation, morphological traits, and/or geographic origin. Within the larger clades, additional structure can be found with subclades (97% average cluster confidence) displaying a second layer of similarity. In some cases, clade structure reflects the relationships among breed varieties."
It revealed many surprises. e.g. in a closely related group, the Am Eskimo dog and Japanese Spitz were created from the German Spitz.
What this analysis does is reveal the history of breeding (mostly via human selection), and results, by crossing within and between clades. e.g. breeds within the terrier clades (breeds sharing terrier traits) as well as between the terriers and the Mastiff clades.
"For instance, there is a long-standing history of terrier and mastiff-type breeds being crossed in the mid-1800s to form multiple bull terrier- and terrier-like breeds such as the Staffordshire Bull Terrier and the Boston Terrier. There is also excessive sharing between the Mastiff clade and the Retriever clade that has not been observed in previous phylogenies, but suggests recent admixture between these breeds or their ancestors. German Shepherd Dogs and related breeds show the largest number of admixture events with independent breeds from multiple clades. German Shepherd Dogs, specifically, have sharing values greater than 95% of background levels with 29 breeds from 13 clades and three of the non-clade breeds. !!Breeds within the German Shepherd clade are the only ones showing significant levels of haplotype sharing with wolves.!!"
All dog breeds involve some level of inbreeding (at individual and population level). That level can be estimated by comparing sets of DNA variants of gene alleles (the two versions of a gene on the same chromosome) that tend to be inherited together (haplotypes). When two alleles are identical, they are said to be homozygous. When different, they are heterozygous. Runs of homozygosity (ROH) in an individual genome results from the inheritance of two copies of an ancestral haplotype in that individual. Thus they are homozygous by descent.
The measure of individual or population level inbreeding is the estimated proportion of a genome that is in ROH. "For all dog breeds, selection has involved some level of inbreeding and this has resulted in a wide range in ROH across breeds." The analyses maps proportions of historical levels of inbreeding of breed groups, breeds and sub-breeds within the groups, and compares them to feral/wild canids, including wolf and coyote.
As you can surmise, selective inbreeding results in high levels of ROH (because humans select for specific behavioral or physical traits and interbreed to retain them). The wolf, coyote, and wild/feral canids, have the lowest level of inbreeding (coyote has the lowest of all, but based on only two individuals).
The study also included analysis of size and breed diversity, as well as structural gene variation. The latter "plays a variety of roles in genome evolution, adaptation, and gene expression." It also searched for signatures of selection among major breed groups.
Of major importance is the analysis of mitochondrial genomes. Keep in mind that mitochondrial genome is inherited only from the female line throughout historical descent. "Across the 1933 individuals, only 887 unique mitochondrial sequences (haplotypes) were observed. The most common was present in 52 individuals, and the 12 most common haplotypes were observed in 20% of samples (393/1933 individuals)." What this means is a limited number of female individuals form the basis of dog breeds. Would be interesting to know how many of those sets are completely or mostly associated with wolf source.
Analyses extended to causal homozygous (two copies of an allele on a chromosome) genotypes for autosomal recessive diseases, risk factors, or traits and their associated genes. As we know, long lines of inbreeding often lead to increased recessive diseases and traits. Interestingly, this information has also been compared to diseases in humans for a few decades, providing insight to diseases in both human and canine.
A conclusion many of us already knew: "German Shepherd Dogs and related breeds show the highest allele sharing with independent breeds from multiple clades."
No comments:
Post a Comment