Hacker News Reader

Comment by axod | Hacker News Reader

Isn't this just fortune telling or astrology for geeks?

Not if you're planning to have kids and want to see if you have any recessive mutations to look out for, like CF or Tay-Sacs.

They also trace your mitochondrial DNA and Y chromosome so you can see where your maternal and paternal lines originated, if you're interested in that sort of thing.

The part you're referring to is probably the disease risk section, where they show you your comparative risk of various diseases compared to the population. This is simply a better resolution version of your family history of disease, which a doctor will ask you if you ever go to one. Are doctors fortune tellers?

To really understand the output of 23andme will take a lot more study and I think that if you're going to let it influence your life's decisions at that level that you should be prepared to spend a substantial amount over and beyond the initial fee here to have the results interpreted by a professional that is able to put them in to the proper perspective.

23andme is expensive infotainment without the required education in both the genetic and the mathematical/statistical background.

I couldn't disagree more. Knowing more about your body is always a good thing -- the fact that many people won't know what to do with that information is inconsequential. It may not make sense for many (most) people to buy into 23andMe, there are a lot of people out there that can understand this, or at least find it interesting to be able to peer into their genetic makeup.

Surely if the average person is told "There's an X chance you'll die. There's a Y chance you'll lose a leg. There's a Z chance you'll be blown up", they'll worry unduly no matter what X,Y,Z are.

I can see what you're saying, that if you're a qualified doctor, or have studied medicine for years, then 23andme could be useful, but for everyone else, I think it could be very dangerous.

Surely those things can be tested for properly by a doctor if you're planning kids and believe you may be high risk.

I don't know how things are in the US (I know health care is 'interesting' over there), but one of the jobs of a doctor is to take information and interpret those results for the patient. That's why they spend years training.

Giving patients access to tests directly doesn't seem a good idea at all to me. It's like showing a windows user a stack trace when a program crashes. Most of them won't have any clue how to interpret that stack trace, they'll probably get over-worried their computer is broken or has a virus, etc etc.

People who fork out money for genetic tests like these aren't windows users. I'm a linux user. I also have undergraduate and graduate degrees in biology, and my SO who also got the test done (and with whom I'm planning to have kids with) makes phylogenies for a living. We're perfectly capable of interpreting this information.

Doctors are a fine filter for stupid people, but it they can be infuriating for people like us. My mother is a biochemical engineer, and when she was pregnant with me she was so frustrated by the doctor's dissembling over her amnio, she grabbed the test results out of his hands and read them for herself.

I'm regularly frustrated with the lack of doctors' honesty. I don't need one to decide what I do and do not need to know about myself. I think we not only do we need more direct-to-consumer medical testing, I think we should have better access to our own medical information.

You would be right if the 23andme tests were accurate, but they really aren't. They only test for the most common mutations for certain genes so they don't replace going to see a clinical geneticist. Also, the vast majority of conditions you could have aren't even tested for at all. A lot of these are the kinds of things where there is only a 1 in 40,000 chance of you having it, but there are so many of them that your chances of having one of those conditions is actually pretty high. But yeah, doctors are terrible at diagnosing chronic illnesses so you really want to figure out what you have on your own and then just go to someone to order the tests for you. But 23andme wouldn't be useful for this unless you get really 'lucky' and have one of the 200 or so things they test for.

Your point on doctors is well taken - I myself hold 2 degrees, one in genetics and a second in computer science. I run whole genome sequencing analysis as part of my job and am acutely aware of the kind of knowledge required to properly interpret these kinds of results. When you consider that many MDs simply have 2 years of college's worth of molecular biology training - well it puts genetics knowledge into perspective.

23andMe = http://en.wikipedia.org/wiki/Single-nucleotide_polymorphism + population genetics

But, it doesn't really give actionable results, especially if you're young and healthy.

In my case it suggested what diet methods are more likely to work, and a disease I should especially watch for in the future.

But mostly this is awesome because the future is here!

The future will be here when we can easily grow new organs, and easily target and switch off cancer cells in case those organs become cancerous.

Most of the stuff listed on the 23andme site is just looking at slight correlations and assuming causation. (Though there are some major correlations that strongly suggest causation.)

Yeah, but the "stuff" on 23andme is the result of years of human genome research. The fact that we can now examine our genetic makeup is a large scientific step forward.

The interpretation performed by 23andMe may be suspect[1], but for the bioinformatics geeks among us the raw data[2] they provide may be useful. I know I'm planning to script a targeted literature scraper that makes use of my SNP data.

1. http://www.gao.gov/products/GAO-10-847T

2. https://www.23andme.com/you/download/

It sounds like you're recapitulating HGMD. You should check that out first before redoing it yourself.

It can be if you want it to be or it can simply be an early adoption of a service that will likely be far more meaningful in the future.

I like it b/c it is helping me learn about genome research in an interesting way, and it's preparing me to be ready to digest the results of a full genome, another few decades of research, etc.

Plus, the genealogy stuff is pretty interesting.

Reply on news.ycombinator.com