Covering the coverage of the fox genome paper

Ah, the beloved Tame Fox Project! I worked in a lab that focused on these foxes for four years, during my PhD in Kukekova Lab at the University of Illinois at Urbana-Champaign. We worked on analyzing the fox genome for publication throughout those four years, and now at last the fox genome paper is published. (You can see my name tucked neatly in the middle of the authors list.)

What does it mean to have the fox genome sequenced and published? There is a flurry of news reporting about it, and I have issues with a lot of that coverage. I'll be covering the coverage here, letting you know what is accurate and what is less so. I'll update this post as I cover more articles.

Feel free to comment with questions, or with news stories you think should be included here.

• Fox ‘tameness gene’ identified in 60-year study (Independent): the title of this story is irresponsible, implying that there is one gene controlling tameness in these foxes and that that gene was discovered. The story itself does a decent job of untangling the facts: that a large number of genes affect tameness, and that the gene that was discovered influences whether the fox wants to continue to interact with humans during a specific behavioral test. However, I wonder how many people just read the headline and took away a very different message.
• A Soviet-era experiment to tame foxes may help reveal genes behind social behavior (Washington Post, Animalia): Great title, intriguing and also accurate. However, the story itself over-emphasizes the morphological difference in the foxes, stating that the Tame Fox Project "spawned an ongoing area of research into how domestication, based purely on behavioral traits, can result in other changes — like curlier tails and changes to fur color." We still don't know if the morphological changes in the tame foxes (which are much less frequent than most journalists suggest) are related to their behavioral changes, or if they're just a result of founder effect. (The lab that produced the current study is betting on founder effect.) The rest of this article is good, with an excellent description of the study's design.
• Sequenced fox genome hints at genetic basis of behavior (ScienceDaily): "today, with the first-ever publication of the fox genome, scientists will begin to understand the genetic basis of tame and aggressive behaviors" - I think this is overstating. The fox genome is an important tool for working with the genetics of tame foxes, and they are an important model for understanding the genetics of tame behavior. But this isn't the beginning of understanding the genetic basis of tame behavior - either we started that a long time ago, or we haven't really started yet, depending on how you look at it. As with other stories, this story also calls out the finding that there were some changes in aggressive foxes in a region similar to the one associated with Williams syndrome (hyperfriendliness, among other traits) in humans. Which is cool - but don't forget they also found changes in regions associated with autism and schizophrenia, which is also cool! (And which gives more perspective to the fact that a lot of changes were found in a lot of regions, and we don't know what any of them mean yet.) This story has a nice description of the behavioral trait associated with the SorCS1 gene, the one gene that the paper focused on that has changes associated with behavioral differences in tame foxes.
• The first detailed map of red foxes’ DNA may reveal domestication secrets (ScienceNews): wow, I really like this one! Read this one! It does a great job summarizing the paper, it pulls out interesting stuff, and it doesn't ever go overboard in its interpretations.
• Friendly Foxes’ Genes Offer Hints to How Dogs Became Domesticated (New York Times): quite short, so other articles are better bets for you to learn more about the study. However, I want to give a shout-out to this one for 1) not saying anything misleading and 2) explaining that tame foxes aren't great pets - something that can be really valuable to include in stories like this one.

On showing dogs in conformation shows

I’ve been cogitating recently on the statement I’ve seen in a few places that “it makes sense to show a dog in conformation shows before breeding it to make sure a judge has a chance to say that the dog has good or bad conformation.” I just posted this to a breed-specific mailing list in response to that statement, and am curious what y’all think of it:

I think the real question is whether a judge selects a dog based on healthy structure or based on something else. I suspect it varies by judge, but the concern is that, given a ring of dogs all with good structure, the dog with some other flashy attribute will win (thick coat, particular head or ear shape). Then people start breeding for that attribute in order to win. Then that attribute gets valued over good structure. I think the fear that this will happen in any given breed is valid, given what we've seen in other breeds - take the show German Shepherd with its very sloping backline or the tastefully plump show Labrador.

What it comes down to for me is, what is the best way to evaluate healthy structure in a dog before breeding? I don’t think conformation shows are that way. I suggest a) making sure the dog is able to work well and without pain b) giving the dog time to mature to see if it has any structural unsoundness and c) having the dog examined by a veterinary orthopedic specialist. There are plenty of structural issues that are just not going to show up on physical exam (whether performed by vet or by judge), which is why (a) and (b) are so important.

Thoughts from the blogosphere?

Bonus dog photo because every post needs a photo (of a purebred dog out of parents who were never shown in coformation shows, and a mixed breed whose parents were probably not selected with any sort of care at all):

The vegetables of genetics

Today I’m working on revisions for my DNA class at IAABC, which starts Monday, April 24. This will be the second time I’ve offered this class; it’s the first in a series of four classes (which you can take in any order, so this one isn’t required for later classes). And the auditor’s price is still super low to encourage people to take it just for fun.

I’m never sure how to promote this class. Will it offer you direct insights into how to modify behavior? It won’t, of course. It will tell you what DNA is and what genes are and how at a low level DNA differences affect traits. For how to apply this stuff to behavior consulting, you should refer to the fourth class in the series, which is about behavioral genetics.

But while the fourth class has that stuff we all want to know in it, to really understand how all that stuff works you really want to take this first class. Sometimes I think of this one as the vegetables class: you have to eat your veggies before you can have your dessert. But I hope it’s not just because I’m a genetics geek that I do think this class has some fascinating material in its own right. It’s not overcooked frozen peas, it’s heirloom tomatoes from the farmer's market. In later classes I’ll talk about the weird ways our DNA can affect our personalities, and in order to deeply understand what I mean, you want to know how DNA is put together and how the body reads the genetic code and how things can go wrong.

And by the way, I make sure all of my classes have something in them for everyone, so if you are a genetics geek too, come take the class for the optional resources, which have loads of articles with new research findings in which we (surprise!) realize DNA is more complicated than we at first thought, and getting more complicated the closer we look at it.

And if anyone can help me explain how to market this funny little class and explain to people that this really is stuff it’s good to know (for behavior consulting but also just for life in the middle of the Genomics Revolution) then please tell me!

Ruminations of a dog scientist on a 96-well plate

I've been doing a lot of bench work in the laboratory lately. This involves filling the tiny little wells on a plate with my ingredients (sample, reagents, primers) and then inserting the plate into a reader. The machine takes the plate up with whirring sounds that always fascinate me. I know there are little robot arms in there moving the plate into place, and I wish I could watch the process. But as I listen to the robot work, I sometimes think: is this the closest I get to living, moving animals now? How did I get here, so separated from fur and behaviors and emotions?

96 well PCR plate

My long term research goal is to understand the differences in how brains work in dogs who suffer from fear issues compared to resilient dogs who take life's arrows a bit more in stride. I'm doing this by studying gene expression in the brains of foxes who have been bred to be fearless (“tame”) or fearful (and aggressive — those who study them just refer to this line as “aggressive,” though).

My approach is, at the moment at least, deeply reductionist: what are the differences in gene expression in a few brain regions in these two lines of foxes? In other words, does one group make more of a certain kind of gene than the other? My hope is that I’ll be able to make some conclusions about the differences in function in these brain regions between the two lines of foxes, and that what I find will be relevant to fearful dogs. But I find myself burrowing deeper and deeper into learning about very small parts of the brain, and then very specific functions of those parts to the exclusion of other parts. Currently I’m learning about the pituitary gland — no, wait, just a particular cell type in the pituitary gland, the corticotroph — no, wait, just a particular set of processes of the corticotroph, how it releases one particular hormone into the bloodstream.

So in my daily work, I do things like take some tissue and extract all the RNA from it (throwing out DNA, proteins, cell structure, all sorts of interesting information — that's not what I'm working on or able to assess at the moment). I use PCR to extract a tiny piece of RNA from the complete transcriptome (all the RNA from that tissue), throwing out even more information. And then assess the expression level of that RNA, resulting in just one number. One number out of all that information after a day’s work.

Behavior can’t really be fully understood using this reductionist approach. If I do find a few important gene expression differences in a few small brain regions, they won’t explain the whole story of why an animal has a fearful personality. They’ll be a tiny, tiny piece of a complicated network of interactions involving genetics and life experience. But in order to get at that tapestry we have to first be able to visualize the threads that make it up. So here I am, in the trenches, doing that.

A recovering shy dog.

From the genetics of dog breeds to stress and reproduction

The other morning I was talking to my husband in bed in an attempt to help him wake up.

Me: So I ran into our friend who walks those three goldens separately yesterday and we had a nice conversation. She said she’d read my blog and had a dog genetics question for me.

Him: mmmppphh

Me: She said she’d heard that 1% of dog genes account for all the differences between breeds and asked me if it was true. I pointed out that 1% of 20,000 is still a lot of genes, and also explained that it's really hard to use statistics like that to describe genomic differences, because you can measure those differences in so many different ways.

Him: Did you tell her that humans and chips are 98% similar genetically?

Me: Yes I did.

Him: But I’ve been seeing that for at least 10, maybe 20 years. Is it still true?

I consulted the internet on my phone.

Me: Let's see... The Smithsonian Institute says we're 1.2% different from them. I think I'll skip this link to the Institute for Creation Research -- is that really the second hit on “human chimp genetic similarity”?! Ah, Wikipedia gives more information: “The alignable sequences within genomes of humans and chimpanzees differ by about 35 million single-nucleotide substitutions. Additionally about 3% of the complete genomes differ by deletions, insertions and duplications. Since mutation rate is relatively constant, roughly one half of these changes occurred in the human lineage.” Well, that’s not true.

Him: What?

Me: Mutation rate isn’t constant.

Him: It’s not?

Me: Well it is closer to constant in specific areas, like parts of the mitochondrial DNA, which we like to use as clocks. But over the whole genome, which is what they're talking about here, no. Different areas evolve at different rates. There are hotspots that go faster. And then the whole species might change faster when its environment suddenly changes. Like if you're in a lovely sunny valley and you're well adapted to it and then suddenly an Ice Age starts and your valley fills with ice and you suddenly have intense selection pressure to change your coat length and thickness and your diet and things like that. The stress itself can change your mutation rate.

Him: Stress can’t change your mutation rate! How would that even work? If a female is stressed, it’s too late, her eggs are already made.

Me: Her grandkids then? Or only sperm have more mutations? Hmm, that’s good point.

I consult the internet again. I find and discard an article about yeast evolving more quickly under stressful conditions. Yeast don't make eggs or sperm as part of their reproductive process.

Me: Here you go. Flies. Close enough to mammals for you? Stress does cause flies to have offspring with more mutations. It makes sense because if you’re stressed, it means you're probably not well adapted to your environment, so you should do the random shuffle with your kids’ genetics in the hopes that something, anything, different will give them a better shot. Mostly they’ll be worse off, but at that point it’s worth if it a few are better off and can pass on those genes.

Him: But how does it work with female flies having already made their eggs before they’re stressed?

Me: I dunno... Hang on... Here we are. OK, so the researchers mutated the males, their sperm.

The reason the researchers mutated the males has to do with how DNA is fixed in male and female fruit flies. There is almost no DNA repair in sperm. But the egg can repair DNA in any sperm that fertilizes it.

So the researchers were basically asking how much of the mutated DNA from the male could slip through the repair processes in the egg. The answer was that eggs from stressed females let a lot more mutations through.

Why would stressed female eggs not fix DNA as well? Probably because fixing DNA perfectly costs lots of energy. And these stressed females may not have had enough energy to spare.

There are two different kinds of DNA repair out there. The one that fixes the DNA perfectly costs a lot of energy. The other kind gets rid of any gross problems but leaves errors behind. This costs less energy but leads to more mutations.

The idea is that stressed females can't afford to use the perfect DNA repair system. So they use the other one. Their kids survive but they have more mutations.

—Stanford at the Tech, Understanding Genetics

 Me: Oh crap now I’m late to take Jack to physical therapy.

...Kind of makes you wonder about puppies conceived in puppy mills or animals conceived in hoarding situations, doesn’t it? Might they have more mutations than animals conceived in less stressful environments?

Too many pairs: DNA, chromatin, and chromosomes

Some facts:

• Molecules of DNA are double-stranded, each strand a perfect complementary copy of the other.
• Humans have 23 pairs of chromosomes.
• Chromosomes are made of DNA.
• We have two copies of each gene.

Where do these two copies of each gene come from — the complementary DNA? Or the pairs of chromosomes? In fact, what is the relationship between DNA and chromosomes? These questions have proven thorny for students in the past, so I'll try here to describe the two different ways in which genetic information is duplicated in a cell.

1. Complementary copies of DNA strands

Molecules of DNA duplicate the information they carry. Each base in a DNA strand is bonded to its complement on the opposite strand: A bonds to T, C bonds to G. It’s like a backup mechanism: if something happens to one part of the strand, the other half is there with the complement of the information. So that’s the first way in which genetic information is paired. And the important part about this pairing is that the pairs are exact complements of each other, like mirror images: the information is duplicated precisely and does not vary.

Part of a DNA strand, demonstrating complementary bases in matching colors

This double-stranded molecule of DNA, then, is wrapped up tightly around proteins called histones, and this set of spools of DNA and histones all together is a material called chromatin.

Chromatin: light blue/green strands of DNA wrapped around bright blue histones.

2. Two copies of each chromosome

Chromosomes are made out of chromatin. In a lot of the pictures of chromosomes in which you can see the chromatin that makes it up, the chromatin looks sort of like yarn woven into a sweater. That’s a reasonable way to think of chromosomes: big structures (big enough that we can see them with a not-too-powerful microscope) made from this yarn-like chromatin.

Cartoon of chromosome made of yarn-like chromatin. (Image by Magnus Manske at Wikipedia.)

 A particular gene is always on a particular chromosome in a given species. For example, the gene for oxytocin, OXT, is always on chromosome 24 in dogs and chromosome 20 in humans. So your chromosomes are very orderly, each one containing a specific set of genes.

Humans have 23 pairs of chromosomes. Dogs have 39 pairs. In fact, Wikipedia has a whole page devoted to the number of pairs of chromosomes in different species. We think of chromosomes as looking like big X. The X is actually the two separate chromosomes in a pair, stuck together during the process of cell division. Usually those two arms of the X are separate in the cell.

Image by JWSchmidt at Wikipedia

In a pair of chromosomes, one chromosome is made of chromatin from one (double) strand of DNA and proteins, which you got from your mother; and the other chromosome is made of chromatin from another (double) strand of DNA and proteins, which you got from your father. So you have, for example, two copies of chromosome 20, one from your mother and one from your father.

Human chromosomes: 23 pairs.

These pairs of chromosomes are the second way in which your genetic information is paired. But this is very different from the exact copy pairing of strands of DNA. Your version of the OXT gene on the chromosome you got from your mother may be slightly different from your version of the OXT gene on the chromosome you got from your father. Where her version had a G-C, his version may have an A-T. (Or it may be just the same.)

DNA differences on two different copies of the same chromosome (see the difference highlighted in blue).

So we have two versions of each gene, one on each chromosome in a matched pair. For a particular gene, we may have two identical copies (in which case we are “homozygous” for that gene) or we may have two different versions (in which case we are “heterozygous” for that gene).

So, yes, we essentially have four instances of each gene in each cell. That’s four instances, but a maximum of two versions of the gene (the doubled instances on a double strand of DNA are always identical complements; it’s only when you compare instances between chromosomes that you may see differences).

In the minds of geneticists, it’s the two versions in the pair of chromosomes which really count. That’s the pair that could differ, after all. And that’s why you'll hear that we have “two” copies of each gene, even though the gene is paired both in the DNA and again between chromosomes.

Dog genome ruminations

The other day I was re-reading the original dog genome paper, as you do. This is the paper published in 2005 to accompany the release of the first full dog genome sequence (of a boxer named Tasha) and accompanying annotation (a mapbook of what genes are located where in the very long sequence of bases that is the genome).

You might think that a genome paper wouldn’t be very interesting, because basically the point of it is to say “here is this genome. We published it. It was a lot of work, and it’s done, and now you can use it.” But most groups try to have something interesting to say in their descriptions of a new genome, and this one actually had a lot of interesting stuff about dog genomics in it.

Don’t just take my word for it. It’s open access, so you can read it for yourself.

Lindblad-Toh, Kerstin, et al. “Genome sequence, comparative analysis and haplotype structure of the domestic dog.” Nature 438.7069 (2005): 803-819.

The dog was one of the earlier mammals to be sequenced, so a lot of this paper consists of comparisons between dog and the other sequences we had at the time, human and mouse. We already knew that humans and mice were more closely related than humans and dogs in one sense: they share a most recent common ancestor. This means that as you follow the branches (and tangles) of the tree of life, first you get a branch that divides the most recent common ancestors of human, mouse, dog, and relatives from species like opossum and chicken; then you get a branch that divides the most recent common ancestors of human and mouse and relatives from dog and relatives; and only then do you get a branch that divides the most recent common ancestors of human and mouse. It looks like this:

Tirindelli, Roberto, et al. "From pheromones to behavior." Physiological reviews 89.3 (2009): 921-956. Fig 5

So we’d expect that human and mouse would share more genomic sequence than dog and human, right? Each of those branches in the tree of life represent a point at which one species becomes two, with resulting divergence in genomic sequence. So if the species divergence between humans and mice happened more recently than the species divergence between humans and dogs, then the genomes of humans and mice should look more similar than the genomes of humans and dogs. But it turns out, as this dog genome introductory paper reports, that dog and human share more genomic sequence, more base pairs, than human and mouse do. So how can that be, if humans and mice are closer together on the branches of the tree?

There are several forces contributing to this result, but the one that made me smile was the different rates at which each species reproduces. In the time since humans, mice, and dogs branched off from their shared common ancestor (before humans and mice branched off from their shared common ancestor), mice have had many more generations than humans and dogs. They reproduce so quickly compared to us and dogs that they have more chances to change their genetics from generation to generation. And as a result, while the number of divisions (places where the tree branches) are greater between human and dog than human and mouse, the number of generations of mice between today’s mouse and that last common ancestor of mice and humans and dogs is greater in mice than in dogs or humans. As the paper’s authors put it:

The lineage-specific divergence rates (human < dog < mouse) are probably explained by differences in metabolic rates or generation times, but the relative contributions of these factors remain unclear.

The other way of looking at it is saying that species age at different rates. So while behaviorally modern humans appeared around 50,000 years ago, and dogs appeared arguably 10,000-32,000 years ago, nevertheless the human population is about 4,000 generations old while the dog population is around 9,000 generations old. Because dog generations are shorter.

We created them, but they’re now older than us. Just like how my dog was younger than me when I got him, but aged right past me. Science!

The history of dog breeds: Victorians, genetics, and the importance of diversity

Note: there is an interview with me at the Huffington Post about this story.

Dog breeds are amazing creations. I can own a series of Golden Retrievers and predict with fair accuracy how each of them will look and act. (Look more than act, but the incredible variety of dog personalities is a story for another day.) Unfortunately, I can also predict with fair accuracy what diseases each of those dogs will have, because with the Golden looks and personality come the Golden genetic disorders. As a new dog owner years ago, I thought of these genetic problems as part of the package I was handed when I got a purebred dog: you choose the looks and personality, and you choose the diseases at the same time. But there’s a lot more to the story of how dog breeds came to be saddled with particular genetic disorders than just happenstance: we made choices when we created breeds and we continue to make choices about their health today.

Figure from: Genome-wide SNP and haplotype analyses reveal a rich history underlying dog domestication Nature, Vol. 464, No. 7290. (17 March 2010), pp. 898-902, doi:10.1038/nature08837 by Bridgett M. vonHoldt, John P. Pollinger, Kirk E. Lohmueller, et al.

The beginning of dog breeds
Humans have been breeding dogs for thousands of years, but for most of our history with them, function was more important than appearance. Dogs were bred to work, and beauty was a side effect: long coats were for keeping warm, small size was for chasing tunneling vermin, long legs were for speed.

During the Victorian era, things changed, with the spectacular growth of dog fancy. Suddenly people were breeding and showing dogs for how they looked, not how they worked. Breeds were no longer loose groups of dogs who looked kind of similar and did a particular job; now for the first time purebred dogs had carefully maintained pedigrees. This was the era when the breed books closed, meaning that breeds were suddenly defined as the set of dogs whose ancestors belonged to a select list. If you wanted to make more Golden Retrievers, you could only breed dogs descended from that original list. If you bred in dogs with unknown ancestry, their offspring were considered mutts and could not be competitively shown, even if they looked just like purebreds.

What happens when you take a relatively small set of dogs and use them to breed a much larger number of dogs? It’s like this small set of dogs is marooned on a desert island with no way to bring in new genetic diversity, and their pedigrees are what marooned them. Their descendants will look like them and act like them – and have their genetic diseases. The genetics from those few founders are all that's available to the descendants. When this reduced genetic diversity is severe, it can be a big problem.

Basenjis and Fanconi syndrome
Reduced genetic diversity is severe in the Basenji breed. This breed originated in Africa, but only the Basenjis descended from a small number of dogs imported to Europe in the 1930s are considered purebred. The diversity in this breed was so low in Basenjis in the Western world that in 1990, one in ten Basenjis suffered from exactly the same genetic disorder, a kidney disease called Fanconi syndrome. One or a few of the founding dogs must have had this disease, and it was passed on to their descendants until a large percentage of Basenjis suffered from it.

The solution: bring in new Basenjis from Africa, breed them to the Western Basenjis, and declare that their offspring may be considered purebred, despite a lack of pedigree.  This was done in 1990 and again in 2013, and the effects are still spreading through the Western Basenji population over several generations. (You can read about the trip to the Congo to acquire African Basenjis.)

Dalmatians and urinary tract stones
But what if there isn't an ancestral population waiting to be harvested? Dalmatians also suffer from a genetic kidney disease, in their case stones in their urinary tract caused by high levels of uric acid. It's a painful disease and there was no way to breed out of it: at one point, every single Dalmatian in existence had uric acid levels above normal canine values.

The solution? The Dalmatian Backcross Project, which began in 1973 with the breeding of a Dalmatian to a Pointer. The project husbanded along a line of Low Uric Acid (LUA) Dalmatians, also known as Normal Uric Acid Dalmatians, because what's low in a Dalmatian is normal in any other breed. Puppies from this original Dalmatian/Pointer cross were tested for uric acid levels, and those with normal levels were bred to purebred Dalmatians. This continued generation after generation until a line of Dalmatians had been bred which looked like Dalmatians, not Pointers, but had normal uric acid levels. As of 2011, LUA Dalmatians have been registered with the American Kennel Club and are now considered purebred Dalmatians. It remains to be seen how long the problem of high uric acid levels will remain common in this breed, but at least now there’s a solution in sight.

What we're doing about genetic health in dog breeds
With these success stories, you’d think the problem would be solved. But these are the only two breeds so far to open their breed books to bring in new genetic diversity. [ETA: readers note that a few other breeds have opened their books, including Border collies, Chinooks, Salukis, and Azawakh in the US, and several breeds in Europe. Fantastic!] Both the Dalmatian and the Basenji had easy to diagnose, easy to understand health problems that were also easy to identify with genetic testing: problems controlled by a single gene. Such diseases are relatively unusual. Take the case of the Golden Retriever, who is prone to developing cancer at greatly accelerated rates compared to most other breeds. Cancer is controlled by a lot of genes and is very hard to genetically test for – and therefore hard to breed away from.

While introducing new genetics into Golden Retrievers is very likely to improve the health of the breed, it’s hard to convince breeders to take the leap. As was the case with the Dalmatian Backcross Project, such an undertaking would mean producing dogs that didn't look like Goldens for a few generations. They could still make great pets, but they couldn’t be shown and they probably couldn't be sold for as much money as a purebred. And there’s no guarantee that their descendants could ever be registered as purebreds – the fight to get LUA Dalmatians accepted was long and hard. I use Goldens as an example because I live with one, but many of the breeds we love suffer from low genetic diversity and associated genetic health concerns.

I believe we need as a society to get past our obsession with historical breeds. We can breed for appearance, but that has to take a back seat to breeding for health. We have a model with the Dalmatian Backcross Project. All we need is the will to improve the genetic health of more breeds.

Jack the Pumpkin King, my 14 year old Golden Retriever. He has so far only developed a small tumor, which was easily removed. However, he suffers from skin allergies and epilepsy, both genetic disorders.

What percent nature? What percent nurture?

The Nature versus Nurture debate is over: we no longer ask if genetics governs personality or if environment does. They work together, and it’s hard to pick their effects apart. But surely we can pick their effects apart a little? For example, if a dog trainer is trying to impress upon their students the importance of getting a puppy from a good breeder who takes behavior into account — or conversely, the importance of bringing a new puppy to a puppy class: what should she tell them? 50/50? 60/40? Surely there are some numbers we can cite?

It’s a tough question, but one that researchers have tackled. The concept is called heritability: the measurement of how much of a trait is due to genetic influences, and how much is due to environment.

Human researchers have it easier than dog researchers, because humans sometimes produce identical twins, and twin and adoption studies form the basis of human heritability studies. Some twins are identical (100% identical genetics), some are fraternal (around 50% similar genetics); some are raised in the same home, and some are adopted out and raised separately. You can do some complex math to all of these situations and come out with conclusions about particular traits. Identical twins more similar than fraternal twins for a particular trait? Strong genetic component. Raised together twins more similar than raised apart twins? Strong environmental component.

These studies have given us some numbers: IQ (how someone scores on a particular standardized test) is about 40-50% heritable. Environment does the rest.

Dog studies are harder. Dogs don’t have identical twins. Theoretically, the best way to study the heritability of personality traits in dogs would be to breed parents who do or do not show the trait in question and assess the puppies, then rinse, wash, and repeat for several generations. But this is expensive and somewhat ethically fraught to do in a laboratory, so we fall back on finding populations of dogs whose personality traits have been well measured and whose pedigrees are well known.

How often does that happen? Not very. But there is a test, the Swedish Dog Mentality Assessment (DMA), which is given to a large percentage of dogs in Sweden and some other European countries. Those crazy, overly-responsible Europeans measure their dogs’ personalities before breeding them, to make sure they're breeding stable dogs. Researchers have mined this resource repeatedly to learn more about the heritability of a variety of personality traits.

As lucky as we are to have this resource, it’s not an ideal one. The DMA is a suite of behavioral assessments which are given to a dog on a particular day in a strange environment by a judge who doesn’t know the dog well. Ideally, personality is best measured over time, by someone who knows the animal very well — its owner. And, in fact, every study I read that evaluated heritability of personality using the DMA noted that one of the most important factors was not genetics but the identity of the judge who gave the test. Did some judges tend to judge more severely than others? Did dogs respond differently (more or less fearfully, perhaps) to different judges? Hard to say, but we know that the reliability of the test suffered as a result.

Perhaps more alarmingly, we’re not really sure about the validity of the test, either. What are these assessments actually measuring? They’re measuring the response of a dog to a particular stimulus in a particular situation. Can this response be generalized to a personality trait? If the dog reacts fearfully to a person wearing a sheet over his head so he looks like a ghost, does that mean the dog is fearful or just that this was a particularly surprising experience? The DMA asserts that it measures playfulness, chase-proneness, curiosity/fearlessness, and most interestingly, aggressiveness. But does it? Studies of the validity of behavioral assessments in shelter dogs — a similar situation in which a series of small tests are given to a dog by a stranger in a strange situation — have repeatedly shown that the subtleties of personality are really hard to measure in this way.

Ideally, a personality heritability study would be designed using the canine behavioral assessment and research questionnaire (C-BARQ), a questionnaire which relies on the dog's owner to assess the dog’s personality through 101 questions. This test has been found to be valid and reliable. And the University of Pennsylvania has a database of the results of this test when given to thousands of different dogs. Except... they don’t have the pedigree information for many (or perhaps not for any) of these dogs. So this isn’t a practical solution, either.

So it’s hard, and I don’t really trust the studies that are out there as a result. What do these studies find? Most studies out there use the DMA or tests like it, and find roughly 20%-50% heritability for most personality traits studied. These numbers might be artificially low, though, because the tests may not be testing real traits — behavior that is stable over time.

I was able to find one study using the C-BARQ, which had much higher heritabilities, around 70%-100%. It's a dramatic difference, but I would hesitate to assign the responsibility for that difference entirely to the C-BARQ. This study used a non-random set of samples, selecting aggressive golden retrievers and dogs related to them. With no control set of non-aggressive goldens and unrelated animals, it’s hard to know how to interpret the study’s results.

So what are the real numbers? I still want to wriggle away from an answer. I don’t think we really know. I’d love to see a C-BARQ study using a random sample — maybe by finding pedigrees for dogs already in their database, if that’s possible. Until then, I’ll guess that the real answer falls in the 30%-60% range for most traits. But, in the end, does it really matter? Genetics are important and environment is important. The best genetics can fail in the face of a poor environment, and the best environment can fail in the face of poor genetics. We can’t predict everything about our next dog; we can just do our best to make a good decision, and then provide the best possible environment for whoever comes home with us.

I owe the inspiration for this post to my students in APDT's Canine Behavioral Genetics course, who asked about the balance of nature versus nurture and would not be satisfied with vague answers.

References

• Strandberg E. & Peter Saetre (2005). Direct genetic, maternal and litter effects on behaviour in German shepherd dogs in Sweden, Livestock Production Science, 93 (1) 33-42. DOI: http://dx.doi.org/10.1016/j.livprodsci.2004.11.004
• Liinamo A.E., Peter A.J. Leegwater, Matthijs B.H. Schilder, Johan A.M. van Arendonk & Bernard A. van Oost (2007). Genetic variation in aggression-related traits in Golden Retriever dogs, Applied Animal Behaviour Science, 104 (1-2) 95-106. DOI: http://dx.doi.org/10.1016/j.applanim.2006.04.025