1706: Genetic Testing

Explain xkcd: It's 'cause you're dumb.
Revision as of 21:19, 13 May 2024 by (talk) (Explanation: Ceveril spilling eerors! ;))
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Genetic Testing
Plus, now I know that I have risk factors for elbow dysplasia, heartworm, parvo, and mange.
Title text: Plus, now I know that I have risk factors for elbow dysplasia, heartworm, parvo, and mange.


Cueball has sent a DNA sample to a genetic genealogy company. The implied premise of the comic is that Cueball intended to send his own DNA to one of the several companies that analyze human DNA samples and provide a report as to the genetic history of that person - examples include notable/famous ancestors or relatives, ethnic background, risk factors for certain medical conditions, etc. However, the result that Cueball receives is consistent with a report for a dog pedigree test, breaking down the percentage of certain breeds present in a dog's ancestry. Megan suggests that Cueball has sent his sample to the wrong company. Cueball appears to agree in principle, but (seriously or jokingly - it is unclear) indicates that he intends to hedge his bets and avoid chocolate just in case he actually is, in fact, a dog. Dogs are generally susceptible to poisoning from theobromine, a compound found in chocolates which causes seizures and heart failure in dogs (and many other creatures). Basically, if Cueball really is a dog, then eating chocolate could kill him.

This comic mirrors a real life example of this where a man sent in three DNA samples to a Canadian company (Viaguard Accu-Metrics) claiming to be able to distinguish between Aboriginal (Canadian Indian/Indigenous peoples) tribes. Two of the samples was of his own DNA, however one of the samples was of a dog's DNA. The company claimed that both him and the dog were 20% Aboriginal. It claimed that they were 12% Abenaki and 8% Mohawk. Later, CNN sent in DNA from people who had confirmed ancestry in India and Russia, while other DNA testing companies got their ancestry correct, Viaguard Accu-Metrics also claimed that both the Russian and the Indian had 20% Aboriginal ancestry.

82% or 94% of genes (depending on how you measure it) are shared between humans and dogs. National Geographic erroneously reported that only 5% of human DNA is shared with dogs and mice, which may have misled Randall Munroe. This leads to several possible interpretations of the comic: It is possible (as Cueball suggests in the last panel) that he is, in fact, a dog with excellent human impersonation skills, or that he somehow shares DNA with a dog. It is possible that Cueball mistakenly sent a sample of a dog's DNA (perhaps his own) somehow thinking that is the method of testing his own DNA. Perhaps Cueball submitted his own (human) DNA to a dog pedigree company and their method of testing includes a presumption of dog DNA, and therefore was able to produce this result from Cueball's sample. Or perhaps this comic is a suggestion that some DNA test companies are scams that do not even perform DNA tests, but simply send out arbitrary reports that are not based on any testing.

The title text refers to the fact that certain dog breeds are more or less susceptible to disease. The diseases he mentions, elbow dysplasia, heartworm, parvo virus and mange are several diseases that can end up killing, disfiguring or disabling dogs, but which humans are generally not susceptible to. As noted above, ancestry DNA test results can inform people about their genetic risk factors for disease, either by specifically investigating your own DNA for those risk factors or, more likely (and less costly) by informing of what risk factors are generally prevalent in your ancestry or others people sharing the same ancestry as you.


[Cueball and Megan are talking to each other.]
Cueball: I sent a DNA sample to one of those "Trace your ancestry" projects.
Megan: How legit are those?
Cueball: No idea. I just figured it'd be fun.
Six weeks later...
[Cueball walks towards Megan with a letter in his hand.]
Cueball: Sweet, got my results back.
Megan: Ooh, share!
Ancestry Report
48% Labrador Retriever
35% Beagle
12% Cocker Spaniel
5% Other
[Megan is holding the report.]
Megan: I think you sent your sample to the wrong service.
Cueball: Just in case, I should probably start avoiding chocolate.

comment.png add a comment! ⋅ comment.png add a topic (use sparingly)! ⋅ Icons-mini-action refresh blue.gif refresh comments!


Regarding the title text, elbow dysplasia is something that genetic testing might find a susceptibility to in dogs, but parvo (canine parvovirus) is a viral infection and heartworms and mange are both parasites. I'm only beginning my veterinary studies so it's possible I'm speaking in ignorance but I've never heard of any genetic factors that make one dog more or less susceptible to any of the latter three than another dog. Generally speaking, all dog owners are advised to get their dogs vaccinated against parvo and kept on a heartworm preventive treatment such as ivermectin (Heartgard and other brands). Ie., these are blanket prevention strategies recommended for all dogs and no attempt is typically made to ascertain susceptibility level before recommending these treatments.

Of course, Randall is writing a comic intended for mass consumption and it's possible he wanted to include some canine ailments that would be more commonly recognized rather than just strictly listing congenital ailments of dogs like entropion or brachycephalic syndrome, which might have caused confusion and ruined the joke. --(I don't have an account yet) (talk) (please sign your comments with ~~~~)

There are genetic variations that can make a host more susceptible or resistant to various infectious diseases (see human sickle cell trait vs. malaria), and there are also variations that make prevention strategies less effective. Parvo is devastating to any dog, but Rottweilers, Dobermans, and some spaniels are reported to have higher risk. My personal experience puts pit-type dogs on that list as well, and this may be associated with variations in vaccine efficacy/compliance. Every case I've seen of parvo in a dog with a history of vaccination has been a pit or a rottie, but that's just my experience as a veterinarian, and I don't have a study to back that up. There is a study from Europe that reported parvo outbreaks despite vaccinations in Bernese mountain dogs and dachshunds. Kali (talk) 14:17, 15 July 2016 (UTC)Kali

I think you're all missing the point of the genetic test determining the susceptibility to diseases. While it might be true (or not) that a genetic test would not be useful to test for susceptibility to the diseases listed for a specific dog, the information given by this test (that cueball is a dog and not a man) would clearly change the expected susceptibility of cueball to dog diseases. You would presume that a man is not susceptible to parvo, but if you found out that you were really a dog, you would change the assumption (just like with chocolate). 03:30, 27 July 2016 (UTC)

Now I really wonder if anyone has actually done this before. It would be fun to see the actual results of this. 08:00, 13 July 2016 (UTC)

Just because a disease is infectious doesn't mean there can't be breed dispositions. For example, parvo in GSDs. I tend to see lots of demodectic mange in bull breeds too. This can be due to factors, such as genetic immune deficiency or particular types of skin/hair which can be inherited (talk) (please sign your comments with ~~~~)

Interestingly 5% seems to be the common shared DNA between humans and dogs: http://news.nationalgeographic.com/news/2005/12/1207_051207_dog_genome.html 11:05, 13 July 2016 (UTC)

I was the editor that made the original explanation. I never intended my word to be the final say; if anything, I was expecting the opposite, since I'm not educated at all in those fields. Regarding the link in parvo, I found this study, which starts that Spaniels are the most susceptible to parvo-enteritis. http://www.ncbi.nlm.nih.gov/pubmed/3003015 14:57, 13 July 2016 (UTC)
That effectively puts Pierre Paul Broca in the wrong, from a genetic POV. More specifically his theory that the main difference between humans and primates stem from their understanding of language, something that also relates to Wernickes and Brocas areas of the brain. Obviously animals have language albeit simpler. As for testing the difference between human and animal... with blood tests this is simple, because human blood contains some unique factors. But when it comes to DNA there may be no obvious telltale signs, which would mean only a computer would be able to spot the difference. If that's true it would explain why the lab doesn't run such a test. Todor (talk) 00:42, 14 July 2016 (UTC)

When I came here, I was rather hoping to see a comment on how legitimate these ancestry services really are. Anyone? 21:10, 13 July 2016 (UTC)

I certainly do not know enough about genetics to edit the explanation nor comment substantively, but elsewhere on the web I've seen that humans' DNA is 82% "homologous" with dogs, and 60% with fruit flies! Presumably this reflects the difference between "homologous" and "shared"? As a non-scientist, 5% does 'feel' sorta low. Miamiclay (talk) 02:45, 14 July 2016 (UTC)

I believe by shared they are referring to shared synteny and the amount of colocalization of chromosomes from different species. This is important in comparative genetics which provides important information for evolutionary research. By looking at the differences and similarities between the genes of different species and their locations and arrangements on similar chromosomes they can see evolutionary division and branching changes in related groups of organisms. Eventually this can lead to fairly accurate estimates of their earliest common ancestor, among other things. But, to actually respond to what you said. I think this is simply a misunderstanding of very technical words; shared and homologous are entirely different terms with very specific meanings in the context of genetics. DNA, genes, phenotype, genotype, chromosome, genome, are all vastly different names for specific structures, it's easy to read a study with no knowledge of the terms and misunderstand what it actually says. In fact, most of the time it has little to no meaning outside of the field it's published in. But, as it stands the explanation is fine for understanding the comic. Lackadaisical (talk) 13:16, 27 July 2016 (UTC)

Well now I'm just curious what a 48% lab, 35% beagle, 12% cocker spaniel and 5% other dog would actually look like. 06:29, 14 July 2016 (UTC)

"Cueball clearly does not understand that he needed to send his own sample for testing" How in the fuck is that the "clear" interpretation of the comic? Last panel and title text both say they are his results. 11:30, 14 July 2016 (UTC)

That is not the only wrong assumption in the explanation ATM. Another user pointed out that it might not be *practically* possible to test DNA for which species it belongs to. It is not normal nor practical to do DNA sequencing, on the contrary usually a lab test means a quick PCR, and those are different enough to tell a person apart with a fail-rate of about one in a billion. Assumptions, assumptions, assumptions. 15:25, 14 July 2016 (UTC)
Found on the wikipedia page for DNA barcoding that "mtDNA barcode to assign a species name to an animal will be ambiguous or erroneous some 23% of the time". Also note that this applies to classifying a higher-order species, not like in the comic which sub-species it belongs to. According the the wikipedia page doing so would be largely bogus. 15:43, 14 July 2016 (UTC)
This also answers the question in the first panel about how legit these tests are. Spoiler alert: They are not. 15:57, 14 July 2016 (UTC)

Consider yourself lucky. My blood work has declared me pregnant three times this year as a male. It's gotten to the point that it's one of the ways that my doctors check to see if the results are valid or not. -- 15:21, 14 July 2016 (UTC)

Test your self for prostate cancer. I'm serious. http://thechart.blogs.cnn.com/2012/11/08/home-pregnancy-tests-may-detect-mens-cancer/ 07:17, 15 July 2016 (UTC)

Should it be mentioned, that 48% is a weird number for ancestry? The simplest way (without incest) to reach that (48.4375% to be precise) is having 1/4 (grandparent) + 1/8 (great grand parent, not related to the grandparent mentioned before) + 1/16 (great great grand parent - also unrelated to those before) +1/32 +1/64. --Lupo (talk) 15:20, 2 January 2020 (UTC)