Hacker News Reader

Comment by ececconi | Hacker News Reader

I was so excited when 23andMe dropped their prices to $99 a few years ago and immediately gave myself a 'Christmas Present' and bought a 23andMe kit.

I seriously thought this was the future. I was so excited to see my health information.

It was really cool to see that my spit could confirm that I'm 1/2 asian and 1/2 european. Also, I found it interesting that I have more neanderthal DNA than 99% of people who took the 23andMe test.

Other than that, all the things I've seen are marginally useful at best.

I log into my account once a year or so when I get an email from 23andMe. I understand these things take a lot of time, but I think a lot of the initial users aren't so ready to tell everyone that they need to sign up for a testing kit. I think people were so much more interested in it years ago.

View on HN

  Other than that, all the things I've seen are marginally useful at best.

Part of it is that the FDA successfully sued 23AM from showing things like health risks.

I still see the health risks, like to heart disease, cancer, etc. Might be because I'm grandfathered in?

Correct.

Yes, the restriction is on marketing products to diagnose, treat, etc. Once you've bought the product, the FDA doesn't prohibit the seller from continuing to provide service (and, to the extent that they contracted to provide the service, stopping providing the service could leave them liable to the purchaser.)

You can always export your raw data from 23andMe and plug it into third party (and I guess foreign) services that will give you some health info. Of course you should probably approach the results with some skepticism. I know someone who found out she had a gene that causes an early and very aggressive cancer and now can take pre-emptive measures that quite likely will save her life.

> I know someone who found out she had a gene that causes an early and very aggressive cancer and now can take pre-emptive measures that quite likely will save her life.

Do you mind if I ask what service she used for this?

Not sure what the poster above used but I personally was faced with the same issue. I got my raw data from 23me and used https://www.promethease.com/. I think you can do it for free yourself with the desktop version http://www.snpedia.com/index.php/Promethease/Desktop but it takes longer anyway I found some useful information from it but it's to be take with a grain of salt

I believe she used promethease.com as the other commenter mentioned (which is what I used too). She took the result to the doctor and they did a more thorough test for that specific gene to confirm (which is probably advisable with any serious/negative result before panic).

It is very likely that someone finding this information would have been able to discover it by looking at their family tree. Serious genetic diseases do have a small de novo rate but by and large are already there just being passed down

Wrong.

For one people used to die a lot younger, many genetic defects will only kill an older, weaker human.

For two the environment is different, e.g. many more cancerogens in and around us. Just have a look at the cancer rates. A middle aged man in US today has a 70% chance of getting cancer during his lifetime.

Sorry mate, what?

The op said a friend discovered they had a genetic mutation that lead to an increased risk of cancer. There are several conditions that do this: multiple endocrine neoplasia (types 1 and 2), li fraumeni syndrome (a p53 mutation), familial adenomatous polyposis (apc mutation), hereditary non-polyposis colorectal cancer and a handful of others.

All of these present with strong family histories. In fact, patients with APC mutations often have cancer before their 30s. Yet have been successfully passing on this mutation for generations now.

If you do not have defective copies of the above (or a small subset of mutations of other regulatory and proto-oncogenes) then you have no greater chance of developing cancer than the next person, lifestyle and virus exposure factors not withstanding.

Now I really don't think you have any idea what you're talking about because your two points are at odds with each other: you first state that we are living longer, and defects 'will kill older, weaker humans'. Bullshit. Your next point, that cancer rates are higher now than at some indeterminate point in the past, is also bullshit and out of context: if we are living longer, then of course we are going to get cancer. But this has no bearing on our genetic susceptibility to cancer over a certain age. In fact, there are no known genetic mutations that predispose to cancer later in life; all the syndromes predispose to cancer early.

I can only suppose you are trolling very hard, and hence hiding behind 'anon1mous'

I didn't realize you could get the raw data. I looked a bit on their website and didn't find anything indicating you can. What format(s) will they export?

Not really sure about formatting options but you can get at it here: https://www.23andme.com/you/download/

The FDA put a stop to crowd-DNA research via the 23AM law suit. Prior to the lawsuit 23AM had a large user base completing questionnaires effectively mapping health risks, traits and allergies to genes. Not sure what the Food and Drug administration has to do with my genes other than regulatory overreach.

This is not correct. They put a stop to 23&Me. There is still plenty of crowd-DNA research going on; for example, my genome and some medical information is online via GCG.

http://www.personalgenomes.org

Sorry, I meant PGP not GCG (GCG was an old bioinformatics software).

I too am a PGP participant!

I'm https://my.pgp-hms.org/profile/hu80855C

They claimed it was a medical device in the same way as a glucose meter.

It has to do with the FDA protecting the public from crappy medical devices and drugs.

IF you want to sell a genetic test, you need to prove to the FDA: (1) that your test is accurate and reproducible, (2) that whatever conclusion you draw from the test is scientifically sound.

The issue was that 23&Me was saying "you have this SNP, that means your risk of X is increased by Y%". The test, the result and the conclusion is not based on a rigorous testing. Hell, some of the risk associated with SNPs is coincidental at best.

That said, I'm fine if people want that info anyways, but 23&Me would need to add a disclaimer that says "Your genetic test results and the risks attributed to them may or may not be accurate."

>It has to do with the FDA protecting the public from crappy medical devices and drugs.

Which devices and drugs were 23&Me selling? I thought it was just crowd-sourced information about traits.

Think how ridiculous it would be if the FDA just shut down a crowd-sourced database about trait correlations. But suddenly when a company like 23&Me correlates the two, it becomes pure, concentrated evil that must be subject to regulation.

23&Me was using a medical test to give users data regarding their risks of certain diseases without sufficiently proving to the FDA that that data was accurate.

It's not a stretch to imagine someone acting on that information with potentially dangerous results; taking extra unneeded supplements, having elective surgery, dramatically changing their lifestyle to reduce the risk of some condition, etc.

I'm so happy the FDA was there to prevent this company from doing the things with my genetic information that I asked them to do. Thank god for government.

Well to be fair, you can also ask a snake oil salesman for his snake oil, but there's a reason we regulate these things. The first users might be high information, but once something starts to get popular, people start making bad decisions. You can argue that people can ask for whatever they want (and they can) but whether you can sell anything you want has long precedent, especially in health related areas where misleading information leads to dramatic decisions being made.

And, to be fair, the reason the FDA came into existence was because companies used to sell things that would seriously injure or kill you, and because there were no regulations they wouldn't advise you of the risks. Caveat Emptor doesn't help you if you have no way to become informed in the first place.

Nobody prevented me from having Illumina sequence my genome (they gave me a hard drive with all the files along with medical interpretation).

The difference is that Illumina did things legally (including hiring clinical genetics analysts who pored over the details and made a number of calls about what was significant).

23&Me did not. They were doing something that they would have known was illegal. The FDA explained this to them, and 23&Me blew them off. Failing to respond to an accusation from the government is an amateur move and 23&Me deserved what they got. They only went back and negotiated afterwards.

You're complaining the government was restricting your freedoms but we have more than adequate evidence that 23&Me was acting irresponsibly.

Out of interest, did you use the "TruGenome Predisposition Screen" offering? (Just looked up Illumina: I've been interested, but the 23&Me pricing/privacy concerns squicked me out)

I did Understanding Your Genome http://www.illumina.com/company/events/understand-your-genom... and I'm not 100% sure it's the same thing.

Then I promptly uploaded my genome including the VCF files they used for risk screening, https://my.pgp-hms.org/profile/hu80855C

Anybody on the internet is free to analyze my genome.

Thanks for the link. They seem to offer a number of different products ranging from the $1,000 "Here's your genome" to $10k "Here's your genome and a detailed analysis of what we've found in it."

What's been your experience with analysis? Something that can be picked up, or something that the professional insights you got were far more useful?

My experience was interesting. The uyg clinical geneticists said I had zero genetic predisposition risk! That's a bit surprising it me. II also manually inspected the results and found that most of the risk variants they did see were just genic and I had only one of the snps out of several that are necessary for diseases such as types of cancer.

Personally I found it less than useless, since I can't reasonably conclude I have no risks. I know enough genomics to conclude that beyond a few well understood cases, genomes have little medical predictive ability.

Here is your data in a format for Enlis Genome Personal: https://s3.amazonaws.com/enlisinstallfiles/customerFiles/PG0...

https://www.enlis.com/download.html

> Which devices and drugs were 23&Me selling?

They were marketing their genetic testing kits and related services for diagnosing genetic diseases and risks. That made it a "medical device" within the scope of the Food and Drug Act, which means that they needed to actually have specified level of testing establishing the validity of the things its was being used to assess.

The FDA noticed the violation, notified 23andMe, said, if you want to keep doing this, you need to meet the requirements, and please tell us how you plan to come into compliance. And 23andMe blew them off for quite a while, so the FDA enforced the law.

>That said, I'm fine if people want that info anyways, but 23&Me would need to add a disclaimer that says "Your genetic test results and the risks attributed to them may or may not be accurate."

But they FDA doesn't allow it, even under this condition.

I think it's ridiculous that the FDA considers this okay:

A) "1,2,3-methyl dethyl has been proven in clinical research to make your life suck less. (This statement not evaluated by the FDA blah blah blah)"

But not this:

B) "You have genes X, Y, and Z. The current state of published researched indicates an elevated risk for skin cancer for this genotype."

Was anyone seriously claiming that 23andme was misrepresenting the state of the research? Were they providing or enabling any treatment based on this research without further review by a doctor?

You are correct, the FDA doesn't allow it.

"1,2,3-methyl dethyl has been proven in clinical research to make your life suck less. (This statement not evaluated by the FDA blah blah blah)"

I'm assuming you saw that on a dietary supplement? That would never fly for an approved drug. Take a look at the FDA website to see what kind of promotion they crack down on. There was a ADD drug company that had an ad with a kid studying and the slogan "He can finally get his homework done". The FDA came down hard on them and said "where is the clinical trial showing that academic performance improves?".

As other posters have stated, 23&Me got busted, submitted a 510(k) (approval of medical device) and then when the FDA came back with questions, they failed to reply. You can find it all in the numerous FDA warning letters.[1]

[1]http://www.fda.gov/ICECI/EnforcementActions/WarningLetters/2...

Their disclaimer said: The genotyping services of 23andMe are performed in LabCorp’s CLIA-certified laboratory. The tests have not been cleared or approved by the FDA but have been analytically validated according to CLIA standards. The information on this page is intended for research and educational purposes only, and is not for diagnostic use.

Thanks for pulling that. As I said, I would be fine with it. The FDA isn't.

I think you're missing an important point. The FDA was executing the laws of the united states. 23&Me refused to even respond when FDA sent them a C&D. Then the FDA shut down their work. 23&Me had every chance to negotiate terms with the FDA, but didn't, until they were forced to.

That's not correct. The FDA made them stop doing something they were doing illegally. Their genetic test is a medical device and they didn't have federal approval.

Further, when the FDA informed them they needed approval, they didn't respond. The FDA finally had to invoke the law.

Also, I don't think 23&Me's test results have enough medical value to be used for evaluating health risks.

Further, 23&me's database has serious problems; for exmaple, they still report tongue rolling as a Mendelian trait. That alone shows that they aren't being careful with their analysis.

Also, I don't think 23&Me's test results have enough medical value to be used for evaluating health risks.

For some things they do, like they test for mutations related to Alzheimer's that can drastically alter risk, ranging from "you're probably never going to get it" to "you're probably going to get it, and there's a good chance it'll be much earlier than you would have imagined".

But their marketing material was vastly overstating the value of their results, which is near 0 for most diseases. And it was misleading people into thinking it was more comprehensive than it really was (like, it's not a full genome, and it doesn't include every SNP that we know matters like most BRCA mutations).

Do they actually still show the Alzheimer's information? If so, I'd be very interested in doing that...

They don't show any medical information for new customers, but you can analyze it yourself with stuff like https://www.promethease.com/

All of those diseases that 23&Me reported on have existing single-purpose genetic tests which are far more accurate.

I've done full-genome too. The report (first from Illumina, then from another site I uploaded my VCFs) was medically useless: I read all the top risks, then did additional research, and found that the risks they reported were contradicted by other SNPs (for example, they'll say you are at risk for X, but only if you have Y SNPs and Z SNPs, and if you check Y and Z you don't have it).

In general, you need a genius/expert to analyze these results right now.

> Also, I don't think 23&Me's test results have enough medical value to be used for evaluating health risks.

Insurance companies would have another view.

That's probably the reason why the FDA wants them to get approval before making such evaluations.

The FDA wants them to get approval becaue what 23&Me was doing was illegal and then when the FDA informed them, 23&Me didn't respond (this is all clearly documented!). The FDA has a requirement to enforce the law.

Do they? What evidence do you have of that? Insurance companies typically reimburse for actual genetic tests, and they are legally prevented by GINA from discriminating against you for the test results.

Relevant news report from when the FDA took regulatory action.

http://www.cbsnews.com/news/fda-warns-23andme-tells-genetic-...

I think it's kinda bullshit that there seems to be no limit on the information that the government, Facebook et. al. can gather on me, but somehow I can't be trusted with information about my own genetic code unless it's been vetted by the FDA. It's not like I could go have a medical procedure solely on the basis of this information. Yet supplements and diet pills can make all kinds of bogus claims with impunity.

>I think it's kinda bullshit that there seems to be no limit on the information that the government, Facebook et. al. can gather on me, but somehow I can't be trusted with information about my own genetic code unless it's been vetted by the FDA.

All the FDA did was prevent them from making unsubstantiated medical claims from the results of their test, in the same way that they would for, say, a new pharmaceutical. 23&Me could always go and perform the necessary clinical trials and submit the evidence to prove their results to the FDA. FDA is doing what the FDA was designed to do. What's the problem?

>Yet supplements and diet pills can make all kinds of bogus claims with impunity.

That's because the supplement lobby launched a huge political campaign and got congress to change the law barring the FDA from investigating supplements, after the FDA tried to regulate them.

For those of us who got on board before the fda acted, we still get access to all that information. And op is right, it isn't overly useful.

I know for example that I am not a carrier of cystic fibrosis, and that I have one copy of ApoE4.

But I could have worked that out with a high degree of certainty anyway - I have a first degree relative that died of Alzheimer's and no family history of CF.

It remains to be seen that the information they provide is anything more than a curiosity. In fact, it could be dangerous for the uninformed to receive this information: prominent on my 'health risks' page is that I have a 57% risk of cardiac disease, with a big red bar next to it representing that it will likely claim me one day.

What it fails to mention is that a white male has a high risk of cardiac disease anyway, and putting a giant red bar next to the words cardiac disease without this context is not likely to be confidence inspiring

FDA is US only. 23AM has a lot of customers from the rest of the world, that get all the health analysis.

So why not shift operations to a less restrictive country?

Make people tick a box that says "I am currently outside the United States" to receive the fancy information, if necessary.

I enjoyed my 23&me account in general. The one piece of info that I've found 'useful' was my eyecolor genotype. I have brown eyes and my husband has blue eyes. But I have blue and brown alleles, so before we had kids we were able to intelligently guess about their eye color.

(It wasn't that useful, but we enjoyed it. We now have one child with blue eyes, and one with brown eyes)

Same here, we were super enthusiastic over their service. On top of it, some of the 'signals' they gave us are clearly wrong information. So, nothing.

We did pay a little more for our tests, but I am not sorry for it, I just wished it worked better.

If you haven't already, check out http://www.snpedia.com/index.php/SNPedia.

It's kind of cool to cross reference your SNPs to NIH studies.

Reply on news.ycombinator.com