Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Home DNA test kits

Tags:
  1. Jan 26, 2012 #1
    Has anyone used a home DNA testing service? I was looking at 23andme service. The reviews are pretty good and it's only $100. Could be interesting. What do you experts think? Worthwhile?
     
  2. jcsd
  3. Jan 26, 2012 #2

    Monique

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    In my genetics department we regularly review literature on studies where disease risk is predicted based on genetic defects. Long term follow-up studies are still lacking, but it is clear that commercial kits can not full-fill their promise.

    If you have a 50% increased chance of getting a disease, does that sound scary? What if you knew the disease only has a prevalence of 0.001% in the population, a 50% increase suddenly sounds a lot less pressing right?

    If you have a very serious genetic disease running in your family, go to a family doctor and ask for genetic counseling. Otherwise, live healthy and save that 100 bucks for the future where the tests could become more reliable. That's my opinion. In a few years you can get your entire genome sequenced for that price. Don't expect that to be the holy grail either, things are not that straightforward as we thought at the beginning of the century when the genome sequencing started :wink:
     
  4. Jan 26, 2012 #3
    I agree with Monique based what I know about these tests. However if the test was reliable and it indicated a 50% chance of having a gene for a disease, the relevant probabilities are 50% and the probability of developing the disease, given you have the gene. I realize that Monique was talking about a 50% increase in the risk, which would not necessarily involve a specific gene, but a predisposition due to other risk factors.
     
    Last edited: Jan 26, 2012
  5. Jan 26, 2012 #4

    Ryan_m_b

    User Avatar

    Staff: Mentor

    I'm not convinced by the utility of getting your DNA sequenced, if it is needed for medical reasons it can be done anyway like Monique says.
     
  6. Jan 26, 2012 #5

    Pythagorean

    User Avatar
    Gold Member

    I would get a DNA test for fun, not for medical

    besides, medicine is just a theory : )
     
  7. Jan 26, 2012 #6
    Thanks for the advice! I would be doing it just for kicks! Maybe I will wait a little longer in order to get more clear results.
     
  8. Jan 27, 2012 #7

    Monique

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I'm not sure if I'm following you there, but most of the tests use numbers that are not easy to interpret by the person receiving the results.

    Also, often multifactorial diseases such as "cardiovascular disease" are mentioned: because they are most prevalent in populations. The pitfall is determining the risk there is not easy and different companies who provide the test will give different results based on the genes tested.
    The gene mutations that are known to cause disease with high penetrance are usually the rare ones, as said: you'll probably know whether they exist in your family or not.

    Another pitfall is that these tests often don't look at mutations, but at certain variants that are associated to a disease: the underlying cause is unknown. Whether the variant is predictive depends on your genetic background (e.g. asian or european decent), this is not taken into account in the tests, even further making them unreliable.

    A big effort has been underway to gain more insight in the meaning of a person's DNA, I'm sure the future will be more informative. Our department will be implementing a version of genome sequencing for diagnostics soon, an interesting development. I doubt the patients get to take their sequence home though.
     
  9. Jan 27, 2012 #8
    Going back to your first post, you indicated that a 50% increase in the risk for a disease has be considered in terms of the prevalence of the disease. So if the prevalence is 0 0010 of the population, than a 50% increase the risk based on the prevalence would indicate a risk of only 0.0015. This is true. I was only saying if having specific gene was associated with a known risk for a disease, then the presence or absence of that gene is the relevant factor, not the prevalence of the gene or disease in the population.
     
  10. Jan 27, 2012 #9

    Pythagorean

    User Avatar
    Gold Member

    I'm curious... do these tests tell you about expressing or just having?
     
  11. Jan 27, 2012 #10

    Ygggdrasil

    User Avatar
    Science Advisor

    The types of DNA tests that these commercial services use look only at whether certain gene variants are present in an individual's DNA. The tests provide no information as to whether that particular gene variant is expressed. Figuring out whether the gene variant is expressed is a tricky question, however, because the answer to that question will vary depending on what cell type in the body you test.
     
  12. Jan 28, 2012 #11

    Monique

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I mentioned that for those tests often causative mutations are not taken into account, but variants/markers that are associated to a disease. I often challenge the geneticists to perform functional studies, but that takes times.

    Ok, but if you know the result of a test then you don't know whether you will get the disease or not (except for Mendelian diseases with 100% penetrance). So it always needs to be correlated to a population prevalence, some people carry risk genes that increase their chance of getting a disease. If the tests tells you that you have a variant that increases your risk of colon cancer, I think you would want to know what that risk then is.

    The test will likely say you have a % increased risk, it probably won't say you have a % risk. Maybe Greg should give us feedback if he takes the test, how the results are given :smile: You need to be aware that relevant risk is more often used than absolute risk, because it sounds more dramatic and thus sells better. I guess the same reason that people often use SEM as an error bar, instead of 95%CI.
     
    Last edited: Jan 28, 2012
  13. Jan 28, 2012 #12
    For the most part, this is what I was talking about. There's a short list of genetic diseases where the defect type and location are known, and a much longer list of genetic diseases where the specific defect, location or both are not known. Collectively, they represent a significant burden of disease in many populations. Since most are likely autosomal recessive, it may be useful to test potential parents to assess disease risk in any children they might have when such testing is reliable and inexpensive. In this case, prevalence, disease severity, treatment options for the child and family history are all considerations. However for the affected individual, the prevalence in the population doesn't matter. For any reliable test which places the individual in a higher risk subgroup, it is the disease prevalence in that subgroup that is relevant.

    http://en.wikipedia.org/wiki/List_of_genetic_disorders
     
    Last edited: Jan 28, 2012
  14. Jan 28, 2012 #13

    bobze

    User Avatar
    Science Advisor
    Gold Member

    It might be fun, yet I wouldn't consider it an informative endeavor. As Monique points out, most of the diseases we (in the West are most worried about), like coronary artery disease or metabolic syndrome, are so convoluted through multifactorial risks that just knowing your "predisposition" (if one could even go so far) wouldn't be all that helpful.

    You'd be better off just reading a list of the top 20 ways people in the US die then altering your life style based on modifiable risk factors associated with those causes of mortality. For example, smoking.

    SW has a good point though in that it could be useful for (as Monique puts it) 100% penetrate disease--Like those that follow simple Mendelian inheritance. I think if you have one of these diseases, you likely already know--So in that regard it might not be so useful. But are you a carrier for cystic fibrosis? I don't know if I am (though a negative family history for both my wife and I, and 3 healthy children would be a good indicator I am not), but some of these companies will tell you if you are (though with CF there is an ever increasing amount of mutations associated with the disease alleles, and I doubt any of these companies test for more than the top 5-10).

    I think a "more fun" DNA test might be the ones that "trace" ancestry through population based SNPs. That might be more worth your money for fun's own sake, in my opinion.
     
  15. Jan 28, 2012 #14
    Thanks all! Bob 23andme also does the ancestor stats too. Apparently good enough to find up to 5th cousins
     
  16. Jan 29, 2012 #15

    apeiron

    User Avatar
    Gold Member

    I got a free test a couple of years back because I was doing an article on the whole deal.

    I'd say do it for fun. It could also potentially pick up something useful. I got no particular surprises except an increased risk of haemochromatosis. The sort of condition you would never know about normally, and is good to be alerted to just on the off-chance.

    The ancestor thing is also fun. Trace your mom and dad back before the last ice age. You will definitely find a lot of 5th cousins with no conceivable connection, but fun to chat to.

    Generally I think the site handles the science side of their findings very well. At $100, its easily worth the admission price.
     
  17. Jan 29, 2012 #16

    lisab

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I'd definitely do it, but I have more personal reasons. See, my dad was raised Mormon, and they are very good at tracing ancestors. Thanks to their work I have a very nice family tree all worked out. But curiously, they have little information about my dad's mother.

    Looking at pictures of her as an elderly woman, she looks like she may have African heritage. A test like that would pick that up, I think.
     
  18. Jan 30, 2012 #17

    Borek

    User Avatar

    Staff: Mentor

    Be ready for some surprises.

    The more DNA testing, the more divorces.
     
  19. Jan 30, 2012 #18
    Haha good one :)
     
  20. Jan 30, 2012 #19

    Monique

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I was doing genetic testing a few years ago in the States, of families volunteering to participate in a scientific study. I sometimes ran into families where it was clear that the child must've been conceived by another man than the listed father. I always wondered whether any of the parents knew that was the case :uhh:
     
  21. Jan 31, 2012 #20
    I ended up going for it! Just sent in my test tube of spit :) I'll report back in a few weeks with a review!
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Home DNA test kits
  1. DNA test question (Replies: 3)

  2. DNA test for Paternity (Replies: 1)

Loading...