Australian Shepherd Health & Genetics Institute

Australian Shepherd Health & Genetics Institute

Generic selectors
Exact matches only
Search in title
Search in content

Genetics FAQs


Are congenital diseases and defects non-hereditary?

There is a common misconception that congenital conditions are not hereditary.  Some are and some are not.  “Congenital” refers to abnormalities present at birth.  Any birth defect would be congenital, but what that defect is would determine whether it is or isn’t inherited.   Dogs born with the heart defect patent ductus arteriosus have an inherited congenital defect.  Dogs born with problems caused by a toxic exposure to the dam during pregnancy would have non-inherited congenital defects.  Breeders are cautioned not to assume that every birth defect is non-hereditary; many of them are.  Seek the advice of your veterinarian.

How many genes to dogs have?

Scientists once thought that dogs and other mammals had tens of thousands of genes, possibly as many as 100,000.   However, once they started working on the human genome sequence they realized the number was much lower.  In the late 1990s they reduced the estimate to around 40,000.  The more detailed their genome map became, the more they reduced the estimate until they finally determined that humans have around 21,000 genes.  The canine genome sequence, completed in 2005, revealed that our dogs have approximately 19,300 genes.  Most mammals seem to have around 20,000 genes.

Why were early estimates of gene numbers so far off?

Huge numbers of genes were once though necessary to provide the instructions for all the parts and operations necessary to make a mammal function.  Now we know that genes aren’t specialists:  They multi-task.  One gene might be responsible for different things at different times of life or in different body tissues.  Recent study of the epigenome – the system that regulates the genes – has shown that various mechanisms turn genes on and off like a switch or up and down like a rheostat depending on life stage, what type of cell the gene is in, or in reaction to environmental factors.

What are linked genes?

Genes that lie close to each other on the same chromosome become linked when the all chromosomes except the sex chromosomes swap sections prior to dividing in half in the process that leads to the production of sperm and eggs.  Some chromosome areas are more prone to separation during recombination so certain genes almost always remain linked.  There is less linkage of dog genes because they have been divided up into more chromosome packages.   Dogs have 78 chromosomes (39 pair) compared to 46 (23 pair) for humans.  Dogs, humans and most other mammals have about the same number of genes.  The current best estimate is around 20,000.

Aren’t genes several generations back too diluted to worry about?

The idea that genes get “diluted” across the generations is a common misconception.  The genes don’t dilute, but every ancestor on average passes half of whatever genes it had to its offspring – one from each gene pair.  Statistically, a dog will have half of a parent’s genes, a quarter of a grandparent’s, an eighth of a great-grandparent’s and so on.  In actuality it may have more or less than that number but in most cases whatever came down will hover around those amounts.   A dog five generations back statistically only passes on 3.13% of its genes.  Chances are low that your dog has any particular one of that ancestor’s approximately 20 thousand genes; however he almost certainly has a few.  After all, 3.13% of 20 thousand is just over 626 genes.

If a dog appears in a pedigree multiple times, then it is more likely that its genes will be passed on.  If you have enough 3.13%’s from the same dog and maybe a 6.5% for an appearance on the 4th generation the percentage of genes contributed can add up even though that dog doesn’t show up on a 3-generation printed pedigree.  This is a “percentage of ancestry” calculation.  Such calculations do not take in to account the fact that different ancestors that are close kin are likely to be carrying many of the same genes so a line-bred pedigree is apt to have more transmission of the same genes down the generations than would be apparent if you only look at how many times a particular dog shows up.

If a dog has only 50% of each parent’s genes, when you breed full siblings together can you re-creating the original dog? 

Since there are only two individuals (each appearing twice) on the grandparent line it is true that the grandpups will, statistically, have half of the common grandsire’s genes and half of the common granddam’s.  However, during the process in which sperm and eggs are formed, the chromosome pairs split; only one from each pair goes into each germ (reproductive) cell.  That way a complete set is created at the time of fertilization.  Statistically speaking, a dog will have a quarter of any particular grandparent’s genes, an eighth of any particular great-grandparent’s etc.  In reality the number will vary, though most dogs will be close to the statistical probability.  There is no way at present to know exactly how many of any given grandparent’s genes a dog has.

When the chromosome pairs split up in the formation of germ cells, any combination of them might wind up in a particular cell.  If you were to number each member of a pair before the split 1 and 2, you might get all 1s in the germ cell or all 2s, but in most cases you will have a mix of 1s and 2s.  To mix things up even more, before the pairs split they trade segments (recombine) so the half that goes into a germ cell usually includes portions of both the original chromosomes given by the grandparents.

In theory, but unlikely in practice, your pup could inherit 50% of a grandparent’s genes if nothing recombined (extremely unlikely) and all the copies from that grandparent wound up in the same the germ cell (also unlikely.)  If you were to have a double-grandparent you could, again in theory but extremely unlikely in practice, get a pup that has 100% of its genes from one grandparent, with half coming via each of its parents.  However, even then you will not have recreated that grandparent because to do so you would have had to put all the original chromosome pairs from the grandparent together in the puppy.  This would require either no recombination at all or recombination in the parents’ generation that totally reversed that from the grandparents plus each pair would have to have the original #1 and #2 chromosomes, no double 1s or double 2s.  This is very, extremely, terribly unlikely to happen.  If you want to recreate Old Granddad, cloning is much easier.

When cells divide the genes get split in two.  Do half the genes just get discarded?

No.  The short answer is they go into another cell, but it’s a bit more complicated than that.  When most cells divide, all the chromosomes – and therefore all the genes – get duplicated with a complete set going into each daughter cell.  All cells need all their chromosomes to function properly, with one exception:  The germ line cells.  Germ line cells are the ones that produce sperm and eggs.  These can have only one copy of each chromosome because they have to pair up with a sperm or egg from the other parent with its own half-set at the time of fertilization. Nothing gets discarded because the chromosomes and their genes which didn’t wind up in one sperm or egg will wind up in another, though only a half-set is going to wind up in any single offspring.

Before chromosome pairs go their separate ways into sperm or eggs they swap parts in a process called recombination.  Therefore every offspring of a particular dog will have a somewhat different set of its parent’s genes.  When there are multiple offspring, as is almost always the case in dogs, most if not all of the genes a dog has are going to get passed along.

Aren’t most health issues caused by doubling up on recessive genes?

Not necessarily.  Most inherited diseases are not caused by a single gene, recessive or otherwise.   Many diseases are polygenic (multi-gene), arising from genes with quantitative effects (throwing things off a little one way or the other,) or due to a combination of genes and environment.  These make up many of the really thorny health issues we struggle with.

Not only are these complexly inherited diseases difficult to manage in a breeding program, they are also difficult to understand.  Many breeders understand basic genetics, but this rapidly-advancing field has become much more complex.   Much of the research today is focused not on finding a single gene, but on determining how genes contribute to development, how they react with other genes, and what regulates there operation in various types of tissue.  The average dog breeder isn’t likely to plow through the full text of published peer-reviewed research, which isn’t exactly easy reading.  It isn’t surprising many have trouble not only understanding the implications of this research but how to apply it to their breeding programs.

Part of ASHGI’s purpose is keeping up-to-date on scientific developments that relate to Aussies and translating “science-speak” into “dog-speak.”

 How many genes in a polygene trait?

There is no set number, but probably more than two or three.  How many for any particular trait had would depend on the trait and how it was defined.  You could say that “blue merle with copper” is a polygene trait (the merle gene, the black/liver gene and two genes that determine tan trim.)  You could also make a good argument for it being a combination of three distinct traits each with only one or two genes:  Color (black), pattern (merle) and tan trim.  A true polygenic trait would not have distinct aspects that may or may not be seen.  In a classic polygenic trait there is a range of phenotypes in a continuum from most to least.

Is there a way to know how much a parent contributes to a polygenic trait?

Not precisely, but observation and record-keeping may help you determine which are the best bet for getting or avoiding a polygenic trait.  A set of specific versions of specific genes is necessary to produce these traits.  The combination a dog has will determine not only what the dog is but what it can produce, which may vary widely.  Hip dysplasia is the classic example canine polygenic disease.  Each parent must contribute something to the mix of genes that produces good hips or bad, but the contribution might not be equal.  Even with the best hip phenotype, some dogs will consistently produce sound hips while others with equally good hips will be more likely to produce HD.  Dogs which produce multiple offspring with HD (or any other polygenic trait), especially with multiple and unrelated mates, carry more “load” for HD – they have more of the gene versions necessary to produce HD even though they themselves have a superior phenotype.  If you are interested in the potential of a dog that has not been bred, has only a few offspring, or whose offspring are too young to know if they exhibit the trait, you need to do pedigree research.  The more connections the dog has to the trait, whether wanted or unwanted, the more likely the dog is to produce the trait.  The IDASH pedigree analysis program is a method owners can use to determine risk for producing diseases.  Pedigree health research can be done via the IDASH Open Health Database.

What is complex inheritance?

If a trait has complex inheritance there are multiple factors influencing the phenotype you observe in a dog.  Those factors may be multiple genes (polygenic inheritance), genes plus environment, or genes plus molecules generated in a cell to regulate their activity.

What is the distinction between “incomplete dominant” and “dominant with incomplete penetrance”?

If a gene has incomplete dominance, the heterozygote (individual with one dominant and one recessive gene) has a phenotype (what you see or can measure) between the dominant and recessive types.  Pelger-Huet Anomaly is an incomplete dominant.  Dogs with two of the defective alleles will die before birth.  Dogs with only one copy will be healthy, but there will be minor abnormalities in some of their blood cells that indicate they carry the gene.  Dogs with two of the normal versions of this gene do not have these blood cell abnormalities.

Incomplete penetrance means that sometimes you see the trait and sometimes you don’t.  The dog could have one or two copies of the responsible allele but never have the disease.  The gene HSF4, which contributes to most hereditary cataracts in Aussies, is incompletely penetrant.  Dogs that do not have the dominant mutation are unlikely to have cataracts, those with one or two copies might, but might not.  As genetic mechanisms are becoming better understood these genes are now usually called risk factor genes.  Having a particular version makes you more likely to have the disease.  How likely will depend on the level of penetrance of that gene version, with high penetrance making disease more likely and low penetrance less so.  When a risk factor gene has been identified and studied, penetrance is expressed as being x (number) or (number)-fold. (Examples:  17-fold or x17)

How much do the grandparents contribute to a litter?

Statistically, each grandparent will contribute 25% of its genes to its grand puppies, great grandparents would provide 13.5%, and so forth as you go back in generations.  However, these are just statistical norms; the actual number will be around the percentage in most cases but can vary.

 Are the dogs in the male and female tail-lines more important than other dogs in the pedigree?

The only things guaranteed to come directly down the female tail-line are the mitochondria, small energy plants within the cells that have their own DNA.  Mitochondria are important to cell function but do not determine the traits breeders select for.

The only thing that comes down the male tail-line is the Y chromosome a sire passes to his sons but never to his daughters.  The tiny Y chromosome carries only a very few genes, all of which are for specifically male traits.

 Are genetic traits passed to 50% of the offspring?

While half of a dogs’ genes come from each parent, trait inheritance is not as simple as that.  Traits may be inherited in several way.  Only simple dominant inheritance will go every time from the parent that has it to the offspring.  Black is an example. There must be at least one black parent (blue merles are black with the merle pattern) or none of the puppies can be black.  Simple recessive traits like liver (red merle is also just a pattern on a liver dog) will not appear in the puppies unless the other parent also has a liver version of the gene.

More complex forms of inheritance – those that require multiple genes or which are environmentally influenced – regularly skip generations and can’t be easily predicted.

Has anybody done any genetic research on Aussie origins?

There have been three multi-breed studies that included Aussies.  The first was associated with the sequencing of the canine genome.  The researchers looked at 85 breeds to see if they could find unique DNA sequences.  They were able to group those breeds into four clades – groups of related breeds.  For most they were able to find sequences that, in a blind test, allowed them to identify the breed.  Aussies were one of four breeds they could not classify in blind tests nor did it clearly fit in any of the clades, probably because of the very early state of canine genomic science at that time coupled with the fact the breed is relatively recent in origin and has a very broad founder base.

A different study looked at Y chromosome variations.  The Y chromosome only passes from sire to son and therefore doesn’t recombine, or swap parts, like the other chromosomes do but over many generations it will accumulate mutations.  Studying the Y chromosomes in a large

 Has genetic research established relationship between Aussies and the Pyrenean Shepherd?

No.  But at this point (2013) no one has done genomic heritage studies that include the Pyrenean Shepherd.  Until a study is done that includes both Aussies and Pyrenean Shepherds we can’t know how much they share.  There are several herding breeds in Europe, many not well known, which share considerable similarity with the original Aussies. European pastoral paintings of centuries past sometimes contain dogs very similar in appearance to Aussies.  All of these present breeds and those seen in old paintings including the Pyrenean Shepherd may derive either from British dogs brought to the Continent or from some as yet unidentified ancestor of both British and Continental herding dogs.

Do purebreds or mix-breeds have more genetic disease?   Is that why designer dogs are healthier?

Mix-breeds, as a class, have more inherited diseases than any single pure breed because of their broader gene pool.  However, the frequency of any given disease is likely to be lower in mix-breeds because the population is more diverse.  The increased frequency of specific genetic issues in specific pure breed’s, over that of the species as a whole, is what makes purebreds good subjects for genetic and biomedical research.

“Designer” dogs – deliberate crosses between breeds – are said to have “hybrid vigor” because by combining two breeds you will not double up on problem genes from either pure breed.  However, not all designer dogs are truly hybrid; because of their popularity some who breed them are going beyond the F1 (first) generation.   If an X-doodle is bred to another X-doodle the puppies will begin to show traits, including diseases, typical of one or the pure breeds.  For instance, if one of those pure breeds often had hip dysplasia, the first generation would be clear but when the true hybrids were bred to each other, the next generation might develop hip dysplasia.

 If pure breeds average around 12.5% COI, the equivalent of half-brothers and sisters, is it really possible to make an outcross?

The average COI for the Australian Shepherd is around 13.5%, slightly higher than the COI for a mating with one common grandparent on both sides (12.5%).  But this is an average and any individual Aussies may have a COI that is much higher or much lower, though most will be around that point.

The only way to know how close two dogs might actually be is to run a COI calculation on a prospective mating between them.  If the result is lower than breed average, they are less related than most dogs in the breed, if it is higher they are not.  The degree of variation from the average will indicate how much or how little related they are.  If there are no common names in 5 generations, the COI is below breed average and is also lower than either parents’ COI you could consider it an outcross.

However, nobody should make achieving the least possible degree of relationship between parents their main goal.  Reducing the level of inbreeding is important, but you need to do it in the larger context of maintaining or improving structural, behavioral, and health traits.

What’s wrong with inbreeding?  Don’t all breeds get established that way?

To some degree, yes.  Breeds have been established in a variety of ways.  Some were distinct regional types developed for various purposes that were more formally bred after the rise of the show system.  Some were developed from a few – often very few – individuals of a particular type imported to Europe from some far corner of the world.  A few came into being through intentional mixing of extant breeds to achieve a particular goal.  In all cases a degree of inbreeding played a part because, once “proper type” was determined the quickest way to make it consistent is through inbreeding, including linebreeding which is a form of “loose” inbreeding.

But once breed type is established it is important to maintain as much genetic diversity within the breed population as possible.  Failure to do so risks inbreeding depression:  Increased health and reproductive issues, and – in the worst case – extinction.  The White English Terrier, derived from the Manchester in the 19th century, contributed to the development of a number of other breeds including the white Bull Terrier.  However, it succumbed to inbreeding depression within 70 years of its creation.

It isn’t that inbreeding is “bad” or “wrong” in and of itself.  The problem arises when it is used too frequently or too stringently.  A breeder may have compelling reason to do a close breeding, however it shouldn’t be done generation after generation.  The overuse of popular sires and the subsequent linebreeding on them and on their progeny and descendants is the major cause of loss of diversity in purebred dogs.

 How can close breeding increase chances of getting unwanted traits?

Every dog has two copies of every gene on every chromosome, with the exception of males and genes on the X and Y chromosomes.  The more closely related the parents are, the more likely it is that they have both got the same version of that gene inherited from a common ancestor.  For example, let’s say you have a litter from a black tri dog and a blue merle bitch who are half-siblings through their sire, a red tri.  Odds are very high that you will be getting some red puppies in the litter thanks to that red tri dog being on both the sire’s and dam’s side.  This is true not only of traits you can see, like coat color, but any other recessive traits that common ancestor might have had or carried.

Not all close breedings are from a half-sibling cross, but the more ancestors that appear on both sides of a pedigree, the more likely your dog is to have inherited two copies of the same version of any given gene.    A coefficient of inbreeding (COI) is a prediction of how likely that is to happen.

Wild species like wolves, coyotes and foxes all look very much alike, even more than all the dogs in a breed sometimes.  Isn’t this an indication that they’re inbred?

Not at all.  There is no animal breeder who selects more ruthlessly than Mother Nature and her selection is not by pedigree, but function.  The uniform look observed in most wild species arises from superb adaptation to fill a particular role in nature.  Thousands of years and countless generations spent living a particular lifestyle, often in a particular type of environment, shapes species both physically and behaviorally into what works best.  What doesn’t work, doesn’t survive.  Where you see consistent significant variation (body size or coloration difference between sexes are two examples) they persist because they contribute to the species’ survival.  Wolves, our dogs’ closest wild relatives, have different coat colors.  White wolves only live in the arctic and black ones in forested regions.  Shaggier coats are found where winters get very cold while shorter coats are ubiquitous in hot desert regions.

When it comes to inbreeding, each species has behaviors which significantly limit its occurrence.  Maturing offspring will leave their mothers or their natal groups to find new territory, usually at sufficient distance that they are unlikely to breed with close relatives.  Animals can move around, but even plants have ways of dispersing their seeds widely or preventing self-pollination to avoid inbreeding.  The most notable exceptions will be species confined to a constricted habitat, like an island, which have no alternative but to cross with relatives

How do they know how much genetic diversity a breed has? 

Where within-breed data is available, the findings were not unexpected.  Breeds with large populations exhibited much more diversity than very rare breeds.  Breeds that had few founders or which had gone through a bottleneck were less diverse than those with many founders and no bottlenecks.

The Aussie has not been included in any of these studies so far (2013).  The breed is of fairly recent origin and has over 300 founders, as determined by a researcher who was looking at breed longevity back in the 1990s.  A founder, in the case of our breed, is a dog with modern day descendants and no known pedigree.  The breed hasn’t suffered any bottlenecks due to things like wars, disease or economic impacts.  The overall diversity in the breed is probably still pretty good.  However, breeding selections – mostly popular sire breeding – have in some cases resulted in fairly high coefficients of inbreeding in some lines.  For example, the show lines–which are largely various offshoots of the old Flintridge line of the 1960s and early 70s with bits of this and that added in–averages about 13.5% COI.  They are effectively all half-siblings.

Other lines vary.  Some have quite high COIs and others are very low.  In the latter case it is probably because the breeders are paying more attention to what the dogs and their near kin are able to accomplish and not trying to create pedigrees that double up on Great Dogs from the recent past.

How can you figure out gene frequency in a breed? 

Gene frequencies can be determined using a formula called Hardy-Weinberg.  It determines frequencies as well as the ratio of genotypes (A/A, A/a and a/a) for a given gene with two versions where one of them is recessive to the other.  When the frequency of a mutation is .2, the number of heterozygotes (one copy) is 32% with 4% recessive homozygotes (two copies) and 64% dominant homozygotes (no copies.)

A population of 50 dogs would have two copies of the gene apiece and therefore 100 amongst them.  At a .2 gene frequency, two of those dogs will have two copies of the recessive allele, for a total of four, and 17 will have only one copy.  17 + 4 = 21 or a frequency of .2 when you round the number off.  This is called Hardy Weinberg Equilibrium.

In real life a number of factors can cause a population to stray from HW equilibrium, however the formula is useful to researchers and the medical community who need to determine possible outcomes of a mating between individuals of particular phenotypes, how much of the population may have a particular genotype, or what the allele frequency is if the percentage of a particular genotype is known.

For instance, if we know 2% of the breed has a recessive trait using HW we can determine that 24% will be heterozygous carriers and the balance homozygous normal.  Gene frequency for the mutation in this case would be .14.

If 4% of Aussies have an inherited disease, how many are carriers?

That would depend on the mode of inheritance for the disease.  If it were a simple dominant there would be no carriers; every affected dog would have to have at least one affected parent.  Based on a single gene recessive mode of inheritance approximately 32% of the population will be carrying the mutation (determined by using the Hardy-Weinberg formula).

But many inherited diseases are not simple single-gene traits.  Predicting exact or even a close approximation of the number of carriers is difficult to impossible if a disease involves multiple genes, though the number of carriers will be more than you would expect from a recessive.  With multi-gene traits to accurately predict outcome you would have to know all of the following:

  • how many genes are involved
  • which are dominant, recessive, or additive
  • which can mask the action of other genes
  • which only express in the presence of certain versions of other genes
  • are there any environmental factors; if so, how do they impact gene activity

At this point in time it would be very difficult to create a formula to predict how many dogs in a population carry genes for this type of trait.  For the same reason, it is difficult to predict litter outcomes when you breed dogs who have previously produced pups with multi-gene diseases.  The best thing breeders can do when one of their dogs has produced a disease with complex inheritance, is not to breed it to mates that have family history for that disease.  If the dog produces the disease multiple times with different and unrelated mates and the trait is a really bad one, it may be best to withdraw the dog from breeding.

What is mtDNA? 

mtDNA = mitochondrial DNA.  Mitochondria are little structures within the cell that serve as “energy plants,” keeping cell function humming along.  Every cell has a batch of them.  Mitochondria have their own DNA (mtDNA) and it is distinct from the nuclear DNA that we normally think of in terms of inheritance.

Unfertilized eggs have mitochondria.  Sperm do, too, but they lose them at the time of fertilization so only the mother’s mtDNA will be found in the offspring.   The mtDNA does not recombine like nuclear DNA, so it does not change from generation to generation.  It may mutate, but the rate at which it does so is very slow.  A puppy receives its mtDNA from its mother.  It will be identical to that of its maternal grandmother, and her mother, and so on down the female tail-line with the exception of those very rare mutations.

Can mtDNA be useful to dog breeding?

Your dog’s mtDNA won’t tell you anything about most the traits we are concerned with as breeders because the genes for those are in nuclear DNA.  It can’t be used for parentage verification because it would only tell you if a group of alleged siblings were out of the same mother line but it can’t tell you exactly which female from that line is the dam and won’t tell you anything about the sire.

Since mitochondria are important to cell function, in time we may find that they play a role in some disease processes because a cell with defective mitochondria won’t function properly.  At this point little research has been done on mitochondrial diseases in dogs though some have been identified in humans.

Can mitochondrial DNA (mtDNA) be used to prove whether someone crossbred to a another breed?

While mtDNA has been used extensively in evolutionary genetics to determine the branching of the family tree for a number of species, including the dog family, and to pinpoint when the common ancestor of related species lived, it isn’t useful for determining recent breeding practices.

Let’s propose a hypothetical case where someone has crossbred Breed A to Breed B and backcrossed the offspring and all subsequent generations to Breed A.  If a male of breed B was used all descendants would have mtDNA from Breed A because males can’t pass their mtDNA along.

Since mtDNA passes whole from the mother to her offspring it remains relatively constant from one generation to the next.  Variations do develop, but this happens over a long period of time – much longer than the period during which most of our modern breeds developed.  That said it is possible that the particular mtDNA haplotypes (versions) that occur in Breed B may not be found in Breed A (though if the breeds are closely related they might have the same haplotypes.)  If you suspected that a dog of Breed A had a female ancestor from Breed B, it would have to be descended from that ancestor down he female tail line to have the Breed B mtDNA haplotype. But this would only be useful if the particular Breed B bitch that was used had an mtDNA haplotype not fund in Breed A.

While mtDNA is frequently used in research, I am not aware of any canine mtDNA tests available, particularly for the purposes of determining ancestry.  A breed would be much better served by using the already established parentage verification tests on all breeding dogs.  This wouldn’t reveal any crossbreeding in the past but would prevent it in the future.