Hacker News new | past | comments | ask | show | jobs | submit login
How Anne Wojcicki’s 23andMe Will Mine Its Giant DNA Database (forbes.com/sites/bizcarson)
141 points by woodgrainz on June 8, 2019 | hide | past | favorite | 137 comments



DNA sequences should be viewed differently than other personal data, since they are personal not only for you, but also for your children, grandchildren, siblings, parents, etc. When you make a decision to share this data, you are making a decision for them, too. The implications might not be the same for your great-great-grandchildren, but who can tell whether the impact will be better or worse? How will it be possible to use the information in the future, when technology, science, law, politics, and social customs might be quite different?

We often see complaints about big companies selling their users' personal data, but in this case the decision lies with each individual who shares his DNA sequence. Do you believe it is ethical to share your relatives' personal data without their consent?


On a similar process I find disturbing that when a friend decides to use WhatsApp and the decide to share their contacts with (f...ing) Facebook, they give them my name, surname, phone number, and who knows what else is in that contact card.

They don't ask me, I don't approve that, but still Facebook gets my data. Same applies for Viber, Signal, etc. of course.


I agree with you, though Signal does not get access to your contact list, in case anyone is looking for a less user hostile messaging app.

https://support.signal.org/hc/en-us/articles/360007061452-Do...


Using WhatsApp just shares the raw phone number. None of those other details leave your phone.

(Of course sharing /w the Facebook app is probably a completely different story.)


It doesn’t matter, you only need a small % of your contacts to leak their contacts to Facebook/WhatsApp and it’s trivial to build a graph of almost everyone you know


You are leaving a DNA trail in your wake, and it's only private or personal insofar as no one looked at it. But it seems like it's something that we really have no hopes of keeping private over any reasonable timeline into the future.


Aren't we moving the goalposts as technology advances?

For instance, your DNA used to be private in the past simply because no one could do anything with it even if they would get it. So you benefited from this privacy aspect by default. In the next 20 years your DNA will probably be in multiple databases somewhere even if you never offered it to companies.

Similarly, in the past you may have benefited from the privacy of your own home. You could say or do anything you wanted and that would be kept private (for the most part). Now, with all the "always-on" smartphones and smart home devices and surveillance cameras, everything you do or say in your home will be on someone else's server, which can be data mined, sold to third-parties, requested by various law enforcement agencies, and stolen by cybercriminals.

It seems to me that from your point of view and with enough technology advances we'll have 0% privacy in the future. Everything around us will listen to us and watch us, and then others, with who you may have never interacted, will also get to see and analyze all of that data.

So the technology could enable all of this -- but the question is should we let it? Privacy is a human right for a very good reason -- abuses against someone's private life can lead to all sorts of nefarious things against that person, whether it's something as "benign" as increasing your insurance rates to not offering your free/cheap healthcare because you "live too dangerously" or malicious actors and government agents using it to destroy your life for profit or personal vendettas.


In the case of smart devices in your home we do still have a choice -- just don't buy them or use them, or ensure that the ones you do use have strong protections, etc. However with DNA it is like you are screaming it in countless public places daily and it was only ignorance that protected its privacy. That time is rapidly coming to a close. The cost and time of processing it, once prohibited, has fallen through the floor and soon enough just doing mass dragnets is going to be trivial. It seems likely that generating facial profiles from DNA is not too far off either. Couple that with a facial recognition system and you have a deep identity.

We should fight it as possible, but we'll see how far that gets us.


Unless research in this area is policed, outlawing the practice will drive it underground or to another country with less qualms. One problem with trying to ban incorporeal things like genetic sequencing algorithms is that you can't control supply like you can with guns or really any physical good that has a high capital expenditure requirement.

You say "should we let it" like that wouldn't mean getting way more draconian about policing information. How do you prevent people from sharing their own DNA? How do you keep people from reading and remembering it?


Interesting, did anyone ever try to massively analyze DNA by collecting, I don't know, hairs or used cutlery or other waste that contains DNA?


Shoot! Yes! My partner showed me a story about this recently where a person reconstructed potential faces from DNA collected from discarded cigarette butts, hair, and other items.

They used the faces in an art display.

I’ll see if I can find it but I’m not confident—

Edit: found it easier than I thought:

https://www.cnn.com/2013/09/04/tech/innovation/dna-face-scul...

This is a privacy matter, I think. It seems haphazard to allow this kind of thing wantonly...



This is an inevitable future. Adtech and martech companies already pay for customer data. Why would they not set up ways to gather this data?


The premise of the film Gattaca (1997) is slowly coming true.

https://www.imdb.com/title/tt0119177/


It's similar to facial recognition.

It's not about the ability to look at something, face image or DNA. It's about ability to remember that information and use it.

Some argue that they want ownership of information that concern themselves, and right to deny others from handling that without explicit consent. Others argue that collection is alright but sharing (with or without exceptions) isn't. Others disagree completely and say that information belongs to whoever holds it, as it's immoral to deprive person of rights to have memories. It's also arguable that privacy laws may apply differently between persons and companies. Or between natural and artificial memory banks. Or that different rules should apply based on volume of data stored, like number of persons involved. I've read many different opinions on this matter. Lean towards some, but haven't firmly decided on any.

I don't think this will be fully resolved until someone would invent eidetic memory drugs or implants or something like that, and we'll have to erase the current boundary between humans and machines.


If people have that attitude then it will most certainly be the case.


I agree with you. However, I felt important pointing out that we already have something like that happening, where users consent to share their personal data and in the process share data that isn’t really theirs as well. Just to clarify, i do not think that it is supposed to make sharing data of other people in this way any less bad.

Famous example (albeit of a smaller significance, since the data of users shared was public for all their friends) - the whole Facebook/Cambridge Analytica scandal. Users were giving up their data using their survey app, but they ended up sharing the profile info of their friends as well without their permission.


You literally drop your DNA in public everywhere you go. It’d be like walking around dropping copies of your bank statements 24/7 and then saying that they are private. Nearly all of it is a straight clone off the main branch anyway. So I wouldn’t worry about it too much. Source—I do genomics work at a cancer lab.


Right, however in public its a bit more challenging to identify a DNA sample with the identity of its owner. One can get this via a targeted attack, yet this seems in the same spirit of going through someone's trash to find their bank statements.


That’s a fair point. Operating a global database that connects someone’s DNA to their identity brings up different issues than just being able to easily sequence someone’s DNA in a “targeted attack”. I think worrying about keeping your DNA private is a fools errand (See oxford nanopore and extrapolate), but discussing what companies can do with these databases is probably a more fruitful effort.


Anyone surprised? For this and other reasons I've refused to use their service and other services like theirs.

Here's a service. Pay to get your panel/exome/whole genome sequenced, put your data on gcp/aws/azure, user open source to do variant analysis and pay a specialist for their computer aided analysis.

Guess what, you own your data end to end. I imagine the above could be had for less than $2k. Some startup will write the software stack to do this and simply sell licenses for software you run anywhere you want.


Federal law requires that labs retain the results of genetic testing for a minimum of 10 years.

> Reports of genetic testing must be retained for at least 10 years. Electronic records are acceptable. Specific regulations for specimen retention are not proposed, but each laboratory must have a written policy defining its own specimen retention policy.

http://clinchem.aaccjnls.org/content/45/5/739


This is for CLIA tests, i.e. clinical tests. 23andMe is explicitly NOT this as they can't make medical claims. Just getting a piece of DNA sequenced does not require this kind of retention; GP is right, this is a viable business plan.

Data retention is the next big wave and these companies would do well to clue in. There is zero chance they will find anything medically useful with their database and a ton of outstabding liability once people realize how it can be abused.


23andMe is CLIA certified and FDA authorized, and makes many medical claims. From [1]:

"CLIA certification and CAP accreditation 23andMe laboratory testing is done in U.S. laboratories certified to meet CLIA (Clinical Laboratory Improvement Amendments of 1988) standards, including qualifications for individuals performing testing and other standards to ensure the accuracy and reliability of results. The laboratory is also accredited by the College of American Pathologists (CAP), which has served as a model for various federal, state, and private laboratory accreditation programs throughout the world."

FDA authorizations for 23andMe's personal genome service are available online, e.g. [2] for Alzheimer's disease risk reporting based on the E4 variant of the APOE gene.

The company also offers ancestry reports, which are not clinical and thus covered by CLIA. But medical claims in 23andMe's health reports do comply with CLIA and other regulations.

1. https://medical.23andme.com/dna-kits/#clia

2. https://www.accessdata.fda.gov/cdrh_docs/pdf16/DEN160026.pdf


Hmm, news to me. In any case the point still stands that you can do SNP sequencing without being CLIA certified.


No one should be surprised, their database is extremely valuable. They have a labeled database of millions of genomes. Once they sequence them all, then they just need to buy Luminosity (which is basically a site full of IQ tests) and find their common customers. If they have a millions common customers, then they can build a predictive model for IQ. Repeat with any other trait you want to predict. Hell, buy a video game company and embed behavioural tests within it.

Sell it for use in embryo selection, employee profiling, to governments.


Huxley?


There's a lot of churn in that market of startups doing it and stopping doing it.

The scientific community has already created all the tough parts of the software stack, and everything has been commoditized already, there's just not much of a customer base.

The customers that it's really really useful for are those with rare phenotypes that want to learn more, and they can't find a doctor with expertise, and there's no lab specialized in genetic testing for that.

However, those are also the cases where the off the shelf software/sequencing doesn't always give the right answer because it's not a simple small change, and then you need specialists both on the biology and bioinformatics side to really make it useful.


Not surprised, contributing to our understanding of genetics was half the reason I signed up for 23andMe.


I think that's an incredibly small niche. So few people care that much about privacy or have the technical awareness. This solution costs a lot more and requires more work from the user.


Well, once you launch your startup I look forward to being a customer :-)


I'm not a huge fan of GSK making money off this data, but I am a huge fan of having the data be available. Did you know you can download all your 23andMe data and make it publicly available through openhumans.org? Here's mine:

https://www.openhumans.org/member/iandanforth/

I want this data out there to help any and all researchers. The more free, public data exists the easier it is for researchers without GSK levels of cash to make discoveries and contribute back to the community. Just like OSS, someone has to be willing to give away something that has historically been sold. I'm willing to do that and I hope you will as well.


Thank you for being so open. I feel similarly about my own genes but I do not have them sequenced yet. A lot of people are talking about the security concerns and whose making money.

Valid concerns, but I think medicine does need vast amounts of genetic data available to actually make some progress. Medical progress is not advancing fast enough by some very real metrics (5 year cancer survival rate over 40 years is not at ALL impressive when you factor in diagnoses being made earlier).

Ownership of this data is important, but if genetic data was more freely shared I think medical progress would benefit.


>...but I think medicine does need vast amounts of genetic data available to actually make some progress.

How does more genetic data equate to better progress with medicine? Are we testing medicine against specific genome sequences, now? If so, how are we doing that without the source host[s] to test against?

>Medical progress is not advancing fast enough by some very real metrics...

Again, I'm not seeing how these two equate, whatsoever.

For what you're talking about the genomes would have to be reproduced, such as the markers that are the precursors for breastcancer. Then, the tissues, themselves, would have to be reproduced and then you'd have an effective field with which to test new medicines against (unless you just use the people with the markers to test).

How you're getting from a mapped sequence to better medicine is... ...we're simply not there, yet, technologically, I believe. Unless you know something I don't? (Which could very well be the case, admittedly, but I doubt we're at the stage of computer models for genome engineering, tissue growth, cellular division, etc. all in one.)


Large genetic datasets yield medical progress by increasing statistical power of tests. These better tests enable earlier and more targeted treatment.

Take heart disease. It has a significant but complex genetic component. Many genetic variants each contribute a small amount to risk for heart disease. If a given person has many small risk variants, the sum total risk -- often called "polygenic score" -- can be relatively high.

People in the top 8% of polygenic scores had a 3x higher risk for heart disease than the general population [1][2]. Through techniques like polygenic scoring, large genetic datasets enable uniquely early detection of high risk for the world's leading cause of death.

[1] https://www.nature.com/articles/s41588-018-0183-z

[2] https://www.vox.com/science-and-health/2018/8/24/17759772/ge...


Ah, yes, polygenic scoring and the heart disease increase rate. Did you happen to catch this[0] refutation of the single source's work?

[0] - https://twitter.com/cecilejanssens/status/103135930540723404...

Also, you're posing prediction and targeted treatment but you haven't posited how Bob's mapped genome sitting in 23andme will be used for medical treatment.

As we know, genes are not an emphatic, "this will happen to you," but an increase in likelihood; which still doesn't translate to any emphatic treatments from the genes, themselves, yeah?


> refutation of the single source's work

Janssens seems less skeptical of 23andMe's paper on polygenic score for type 2 diabetes [1][2], which -- interestingly -- positively cites the Khera 2018 paper on polygenic score for heart disease that she critiqued. Some researchers are skeptical, but the medical community generally seems to consider polygenic scores promising for tests [3][4].

> you haven't posited how Bob's mapped genome sitting in 23andme will be used for medical treatment.

Early intervention. Polygenic scores could be used for medical treatment by motivating earlier intervention. That could include stronger recommendations for better diet and exercise, closer monitoring programs, or more precise prescriptions. That, in turn, could reduce disease burden.

[1] https://twitter.com/cecilejanssens/status/113707970323438797...

[2] https://permalinks.23andme.com/pdf/23_19-Type2Diabetes_March...

[3] https://twitter.com/EricTopol/status/1129780543434964993

[4] https://journals.plos.org/plosmedicine/article?id=10.1371/jo...


>...could be used for medical treatment by motivating earlier intervention...

and

>...could include stronger recommendations for better diet and exercise, closer monitoring programs, or more precise prescriptions...

and

>...could reduce disease burden...

This is where the problem delineates for me: We're being massively assumptive in moving from "could" to "is" and "will".

I will, generally, concede the could portion but to assert that it is emphatically happening or going to happen is still far from fruition and to label this science as such, just yet, is overreaching and giving false hope where none should really be given because, then, you'll taint it's benefits with the drawbacks.

Remember: Anonymised data (e.g.: 23andme) only allows a survey of what's relatively known or can be inferred from the anonymised dataset.

To arrive at what you're suggesting, it would have to move into a different realm (I believe), like UK BioBank or GEDMatch but, even then, we're still basing things on speculative science - gambles of percentages that aren't, emphatically, true or false but a kind of "maybe, kind of, sort of, in a way, definitely could or defintely could not" muddied waters.

That, to me, is a far stretch from saying that the data in 23andme is - actually - helping medicine; which I believe is what the OC I replied to emphatically said.


Very cool! Early adopters like you who blaze the trail in the face of uncertainty are crucial to making breakthroughs in healthcare.


researchers can get genetic samples/data without it being personally identifiable. No need to dox yourself.


I am for one glad my data is in there. Disease and aging should be our number 1 priority. If your governement and insurance systems are shitty and spy and exclude people, you should fix that instead of putting barriers to science opportunity.


I partially agree with your angle, however I don't understand the jump from how understanding disease and giving my DNA to a for profit company that may not fully be researching, but showcasing disease, removes "barriers to science". Care to expand?


If not for 23andme, who would have created a comparably big database of SNPs? I dont understand what showcasing disease means.

I know we are all good people, but most of novel science happened with methods that we would find totally unacceptable today, from lobotomizing people, to deafferenting cats, to sending dogs to space. If 23andme didnt have established a large user base, it would be impossible to get a comparably big database given today's regulations (and climate).


> I dont understand what showcasing disease means.

What I mean is you pay 23andme money, they showcase your genetics which, as I understand it, include lineage, risk of disease, etc. 23andme isn't akin to a blood panel in my mind. At least not at this point since they sell on other marketing aspects that are far more consumer oriented.

> I know we are all good people, but most of novel science happened with methods that we would find totally unacceptable today, from lobotomizing people, to deafferenting cats, to sending dogs to space. If 23andme didnt have established a large user base, it would be impossible to get a comparably big database given today's regulations (and climate).

I don't agree with this. You're conflating lobotomy and, again, a for profit company that is collecting a massive database of DNA as we speak and all of the use cases are not yet defined. There is a significant difference between scientific research and running a for profit company. If we can't agree the motives between those two things are different then that is the clear disconnect. Swap out your analogy with Google or Facebook. Are we all glad they have established a large user base? I'm not. And just like those companies it may be that you may think 23andme having your data today is a good thing but that could change quickly when your data is sold for different intentions later on. Do you agree you may not always hold that perspective? What if 23andme starts to sell data to insurance companies after they lobby away having to cover preexisting conditions and then buy your user data to deny you coverage?


I actually view 23andme as a way to extract a bunch of SNPs cheaply. Nowadays it is commonplace and cheap but back then it was difficult. They let you download your data, they provide lineage reports, and can help you find distant relatives. Their health reports (risk associated with specific locations) are not allowed in most of europe. The good thing about 23andme is that there are millions of people genotyped now who can contribute their dna to open research databases. Some related companies have sprung up based on these data like GEDmatch , and promethease (which offers health reports even to europeans).

As for the privacy concerns i have a bit different views. I think it s debateable whether DNA is intellectual property of its owner and whether it could be protected as such. Very few humans can claim that their DNA information is the output of their labor, and it is very highly shared among humans, so there should be a discussion to what extent genetic information should be subject to fair use. Having widely available open datasets would reduce the value of these big centralized databases.


You've gone down Conflation Rd. again. DNA being "intellectual property" has nothing to do with privacy in this case. But having access to someone's DNA means their privacy can be more easily compromised. As others have pointed out your DNA also associates your relatives. While I can grasp the good, a DNA treasure trove like 23andme has will be used illegitimately at some point. We've been to this dance before and we know what it looks like when the music stops.


You do realise that there are plenty of others working on this for research? A good example is the UK Biobank, http://www.ukbiobank.ac.uk/about-biobank-uk/


Surely there are but the span seems to be much smaller (500.000 vs 23andme's 10M)

And it 's not like they lock-in your data, you can export them to contribute to other studies if possible.


UK Biobank has a smaller cohort but deeper and more reliable data.

All UK Biobank participants were tested for blood pressure, bone mineral density, grip strength, BMI, etc. The last 200,000 participants underwent detailed tests of cognitive function [1].

23andMe asks customers for health information, but self-reports are not usually as reliable as the clinician-administered tests done in UK Biobank.

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6451771/


I guess it's analogous to the pros/cons of allowing software to send crash reports... Withholding your data protects your privacy and prevents abuse, but it also causes your problems to be underrepresented in a public data set, which makes it less likely they will be addressed.


> Disease and aging should be our number 1 priority

Humans are ballooning as a species and hurting the planet and causing other species to go extinct... and you want to eliminate disease and aging from said species?


Of course agent Smith. What better way to exterminate the wretched human species than to make it die of overpopulation?


Second derivative of population is negative.

I understand the point of view that preserving species is a moral good, but it is just one of many possibilities and could very well be wrong.


Only specific sub groups are doing that:

https://www.populationpyramid.net/static/population-projecti...

Or to be specific, Africa and West Asia, due to a lack of access to contraception. That can be solved by redirecting 100% of aid towards education, contraception and abortion for women, and away from food aid.


Your proposed solution may sound unorthodox, but it's an interesting view on how to solve the problem in the long term. Although I don't think it will be viable politically nor socially accepted.


I, for one, am glad my data is not in there.

Our government and insurance systems are shitty, and until they are fixed, disease and aging won't be our number 1 priority.

Even individually, our DNA is probably worth a bundle to many companies. To me, my DNA is as near priceless as something can get. I'm glad I didn't pay someone to take it.

The profitable, yet dystopian uses of this information are potentially limitless (see the most recent thread about how the Nazis used IBM technology), but the regulations are negligible to non-existent. I'm more than happy to not be one of the victims of these uses for whom laws and regulations are made after-the-fact, if they're made at all.


This definitely underscores the point that government is not too big. If government regulation we’re keeping the markets free and efficient then DNA and personal data would be considered a good or service owned by the individual.


Needing better regulation in this area doesn't seem to me to conflict with making government smaller. If the government is wasting resources (and it is), then it can do both.


I truly only care about the non-renewable resources it over uses. The economy alone will never ensure everyone has enough to live on its own. If that were the case then the US government wouldn’t be the number one, singular employer in the world. (Or whatever the qualification is.)


If this data was hacked, and your genetic information became public, would you feel the same way?


My genetic and biometric information is already in some government's databases, sometimes with more actionable information than "elevated risk of restless legs syndrome". Associating a list of 500K SNPs with my name would bother me but not that much really. It would take a lot of work to make use of it anyway and i cant think of a reason this would harm me gravely(+). Anonymized information of my DNA would bother me even less, as i m not more exceptional genetically than most people. Tbh, this is not my information, i share DNA with most of the species on earth and i didn't really create that information (unlike, say my google searches), nor did my parents. Thats why i oppose to genetic patents too, and think that DNA should be public ___domain.

(+): Well if someone wants to engineer a deadly virus that targets specifically me, then what can i say , that's at least a memorable way to die. But they can do that from a piece of my hair too.


Honest question - why does it matter? Can they get access to my bank account with that info? My email? Can they impersonate me? How is it different than someone taking my picture?

Don’t I leave a trail of DNA everywhere I go anyway, from shedding dead skin cells?


Yes, an adversary could target you and illegally collect your DNA. But does that mean you should make it easy for them and publicize the data?

There are very serious reasons to keep your genetic information private [0]. Unlike a photo, genetic data reveals sensitive medical information and more...

[0] https://youtube.com/watch?v=HKQDSgBHPfY


Not GP, but personally I do not mind. I have voluntarily shared my DNA information with a number of people.

Health insurance is not important here (UK), and life insurance is much less of an important concern to me as I will not have children. But I can see why if that is not the case, you would be more concerned.

I do have some concerns about it being used by the police, though public availability is unlikely to make a difference there. I've never committed any crimes, but the police are known for using scientifically and statistically invalid methods to prosecute.


And if an employer decides to use your genetic information to discriminate against you, how would you know?

Also, in 10-20 years, what if new threats arise?


I am sorry, but I do not understand this.

If it's valid trait that employer (or insurance) finds actually having negative impact - is it really a discrimination (other than you can't really say whether it's expressed).


In the US, it is illegal for an employer or health insurance company to discriminate based on genetics [0]. But if you fail to receive a promotion because your employer found out about your genetic disease, how would you know?

[0] https://en.wikipedia.org/wiki/Genetic_Information_Nondiscrim...


> how would you know

Same way you get to know your KPI's and other assessments?

I guess that act is good after all. No one can tell you've got an actual disease from these tests, even if on big scale the correlation most obviously is there.


Nice fantasy world to live in, where disease an aging are our number one priority and the insurance system is fixable.

Why not fix both of those things first before feeding our personal data into a fundamentally broken system?


I fully agree, but I think that is easier said than done.

It is conceivable that this would require massive amounts of money to lobby politicians and inform the public of an issues that most people don't give a shit about (Hell, people are so scientifically averse that vaccination is still a significant issues in 2019).

What makes this even more challenging is that there are parties whose interest is to exploit that information for their personal gain rather than helping the individuals with that DNA (e.g. insurance companies, government agencies, etc.). Those parties likely have access to more money and are more capable of lobbying and propagandizing for their interests.


No great advancement comes without dangers. But at the moment we seem to be stifling the benefits while allowing government surveillance to go through "for the good of society".


I agree. I wasn't implying that we should make DNA sequencing and storage illegal.


How about we fix it before putting people at massive risk of harm, and risk poisoning the well on genetic research forever?


right, i dont see anyone arguing for that in here though


The main priority of these medical and genetic companies is to make money. Safety, ethics and privacy are secondary until they effect the profit. It seems increasingly apparent that public funding and open source efforts are required to advance the science, healing and education of our planet.

DNA SNPs are one thing, but genetic expression (phenotype) is immensely complex due to the number of variables. For example, mammalian immune systems are highly redundant across the body and continually signal with cytokines and chemokines to manipulate the immune response that is internally/externally environment dependent. Our bodies continually work to remove/detox OR sequester pathogens, toxins, metals, etc. This usually results in isolated and/or systemic inflammation across the body. Many of the pathogens manifest with overlapping symptoms making it hard to isolate without better diagnostic tools. For example, my two strains of Bartonella cause inflammation in many similar areas that Lyme bacteria do and Babesia has another similar set. It is hard to isolate those symptoms from the mold toxins and metal (aluminum, arsenic, lead) buildup that I also have. Functional doctors believe the HLA SNPs play a role in accumulation rather than removal. Enter saunas/sweat tents/etc that many cultures used for centuries (lymphatic detox).

Enter public funding to invent better and cheaper measuring equipment where each of us owns the data. Privacy protection is also important. We live on a planet with immense biodiversity, albeit shrinking daily due to human activities, where we coexist with bacteria, pathogens, molecules, industrial toxins, etc -- It's time we start to learn more about everything ... gathering data in a coherent schema and applying ML can certainly help.

That is why there are a number of studies that list various pathogens in brain deposits for Alzheimers, MS, Giant Cell, etc -- Feel free to google and research many of the terms above and reach your own conclusions :-)


When will people place a value on their personal data? A video uploaded with a copyrighted song in the background can be challenged by the IP owner and often is. But even with public outrage over things like PI breaches at Equifax, 23andme gets full ownership of peoples’ dna.


Let’s play devils advocate - “So what? What is the big deal if my DNA info is out in the public?”


Are you comfortable paying more for health insurance because you lost the generic lottery? How about facing persecution because of your ancestry? Hiring discrimination?

(Of course the data doesn't have to be accurate or the analysis correct to suffer.)


We already have hiring discrimination based on gender, race, intelligence and who knows what else... It turns out many of these traits have a large genetic factor (there’s obviously nature vs. nurture ambiguity here too).

The solution is to ban discrimination. Banning a technology that has the power to actually make the world a better place through better medical interventions is not a trade off I want to make in the name of ending genetic discrimination (I write this as an African American).


It’s pretty hard to claim and prove discrimination. I don’t know how an individual person could defend themselves against genetic discrimination. It’s far easier to endow people with ownership of themselves their data and components. Maybe a law that requires dna data to follow a sort of “disclosure of stewardship.” Ie the dna owner has to be given the ability to track, remove, and submit their data - with no “opt-ins” without prior authorization.


But it DOES costs more to insure you in case you have health issues. Your health issues aren't merely a quirk that happen to be correlated with higher chances of needing insurance (such as what people used to say about red cars). They are literally what give you a higher chance of insurance.

The solution is a single-payer system, preferably state-run health care.


That's been illegal since 2010.

And besides, 23andme's research data is anonymized.


1. Genetic nondiscrimination laws are under constant attack (GINA, the most comprehensive, faced significant weakening this Congressional session - fortunately, it appears to be on the back burner for now). That's in the US. Canada has significantly less genetic privacy legislation, and the Trudeau government is negatively disposed to that which currently exists.

2. There is no such thing as anonymous genetic data once you move past extremely small SNP panels. Even if the data is pooled.


Then how did they identify a murderer from the dna submitted from a cousin/relative?


the user had uploaded their exported 23andme (or similar) data to GEDmatch on their own volition.


You're thinking of the Golden State Killer and GEDmatch.

Although both 23andme and GEDmatch match up relatives, I don't think they share your actual genetic sequence with your matches. 23andme's research and data publications are anonymized.


Health insurance discrimination using genetics illegal (GINA), but life insurance and long-term care insurance is not.


How do you anonymize an individual's DNA dataset (600k-900k DNA markers)?


First of all, anything can have a downside.

Doesn't seem a big enough concern, how likely I'm going to win lottery?

How about the insurance company? Forced them to lose money?

They won't hire me because my ancestry? Well they are stupid then, it's their lost.


How about you are 3% more likely to develop a chronic illness and as a resut no company touches you for anything except manual grunt work where you could be replaced without much trouble?


Which is kinda the plot of the movie Gattaca, where the main character is stuck in low-paying grunt jobs for this reason until he goes to extreme lengths to fake his DNA to get the job he really wants (and is otherwise completely capable of performing well).


Hah, really? Perhaps I should move to Hollywood!


More likely the reason to use all data as much as you can to find the cure.

manual grunt work will be replaced by automation anyway regardless of you develop chronic illness or not.


That is a valid observation and completely orthogonal to the point being made.


Not at all my point. People OWN their DNA. It’s literally theirs. There should be no question about ownership and the rights associated with owning their data. But we’ve got companies effectively declaring permanent ownership, access, and use of peoples DNA without providing the person with any way to opt out of the database at any point-without providing any payment to the individuals most of the time as well. They even know most of your dna if just one of your relatives sends the company their DNA. It does not matter what you perceive as a risk today to this. It’s your ownership over your dna being claimed by a third party.


You don't own your dna. If you have a twin it is shared for one.

You don't own your fingerprint either. Or your height or weight.


Actually, identical twins don't have identical DNA! See the case of Dwayne and Dwight McNair:

https://www.bostonglobe.com/metro/2018/01/12/man-found-guilt...


What a stupid argument.

By the same argument all companies don't own their code or IP as soon as it's in someone else's hands.


Companies retain ownership of their code regardless of where a copy ends up because of a legal framework that defines it such, not by default. DNA, fingerprints, et al. need such framework too.


DNA is just information, sooner or later it will be spread, and leaked. With the advancement of technology, its just getting easier and easier. Its better to just embrace it. Assume all information is public, and from then on focus on solving the issue that arise from it.


This strikes me as destructive apathy.

Child pornography is information that spreads. I believe society is still healthier if that kind of information is never produced or spread in any form.


The difference is dna data mining has the potential to improve health and cure deseseas.


This talk from DEF CON goes into some of the potential risks of sharing DNA. In short, it conceivably opens a person to targeted attacks designed to recognize specific individuals, families, ethnicities, etc.

[0] https://youtube.com/watch?v=HKQDSgBHPfY


You try to get health insurance and provider denies you because you have an allele that makes you susceptible to lung cancer.

Your DNA happens to be at the scene of a crime (you didn't commit). Police do a DNA dragnet. Now good luck getting off that case.


Regarding health insurance, we already prevent insurers from denying coverage based on preexisting condition. Surely we could pass a similar law to cover this case?


And how many times has that protection been challenged by republicans? They’ve voted to repeal it in the republicans controlled house, the senate, filed lawsuits, and the AG stated overturning Obamacare as a goal of the DOJ.


Is this just a "Republicans are the worst" digression, or am I missing the relevance to this thread?


Or worse, the police lab mistakenly identifies you as a DNA match. This happened for years in Houston and people got sent away for a long time before it was fixed


I do not know why the article did not mention it but I think people might feel the case of Henrietta Lacks where there are a lot of thorny questions would be relevant, even though people give their consent to 23andMe now.

Also false positives for DNA matches - overconfidence in results from DNA labs due to popular culture thinking DNA match = 100% guilty combined with numerous controversies over police forensics in recent years, Houston off the top of my head had to throw out every case that their police lab did testing for a range of years due to lab tech incompetence.


You have a credit rating kept by institutions. That's a wellknown issue - false reports, identity theft etc.

Now imagine a health-reporting institution, used by employers etc. Similar issues will happen - mistakes, failures to update. You get a 'high health risk' rating and can never get insurance again? That's an issue.


That more seems like an issue with your health care system than with the fact that your DNA is out there. Some countries don't have this problem.

You leave a DNA trail everywhere you go. There is no way to keep it private.


Wow these responses seem so naïve. There's a difference between "somebody can get your dna (or fingerprints or ___location)", and "anybody anywhere can get a record of anybody's DNA, and they (you) have to trust the middleman to be right". Trust that there's no malicious actors, or hackers, or political trickery. On and on and on.


That’s an issue now, because of invalid data - having real DNA data would only improve things, no?


My friend got it because she wanted to know what the company has to offer with their service and she is okay, with this service using her data I guess.


I got mine done recently and there are so many fun datamining activities.

Same with ubiome.

Do I care that a company can make a profit? No.. I prefer bsd/mit licease as well for this reason.


>Wojcicki was also building out an in-house drug-research group and tapping into a new revenue source: recruiting thousands of patients from 23andMe’s database for pharmaceutical firms’ clinical trials.

This is the most responsible way to use the data.

The way this model works is that a pharmaceutical companies have a question that requires a genetic database to answer. They then pay 23andMe for the results of these tests.

23andMe does not have to give up any customer data to any third party and can at the same time build a solid business model around using their data for positive medical advancements.


Hasn't anyone ever watched Gattaca before shipping their DNA sample to a Silicon Valley company ?


Gattaca is not a documentary


Neither is the 2059 Genetic Equality Program, described in this op-ed: https://www.nytimes.com/2019/05/27/opinion/ted-chiang-future...


I regret taking their test with 200 bucks now. It used to be novel and years back, the concept of a totalitarian society with surveillance, database, social credit was still very remote


In the UK credit scores have been tied to social behaviors like voting. Similarly London has been famed for their CCTV system that allegedly keeps crime low. And let’s not forget the CompStat database system that is credited with turning NYC from one of the most dangerous western cities to one of the safest. My point is, that at least within the time range of cheap genetic tests, these “totalitarian” systems have already been in place.


> voting

How?


Anne Wojcicki, ex-wife of billionaire Sergey Brinand whose sister is CEO of Youtube is at the helm of the most invasive and extensive data set known to man. Don't fool yourselves: DNA databases form the bedrock of de-anonymization so crucial to the Surveillance State, containing the non-consentiory grounding of your existence and your lineage. And let us not forget the easiness of collection, for a sample of sputum suffices.


This is so rude and self-aggrandizing: "Anne Wojcicki is 45 minutes late, something so encoded in her habits as 23andMe’s CEO that employees have stopped complaining about it. They know it’s hardwired. They know it’s hardwired. On a Thursday morning in April, her team is waiting patiently as she swirls into the company’s headquarters"

From the opening sentence, I already don't like her.


Ok, but please don't post unsubstantive comments here.


Why isn't this protected medical information?


Some time ago a friend who works at a biotech startup in the DNA sequencing space told me there’s no way 23andMe’s price covers the sequencing cost and that they must be subsidizing the kits so they can sell the mined data later.


quick america! fix your health system before the insurance companies know all


It's been illegal to discriminate health insurance based on DNA content since 2010


And as we all know, insurance companies never do illegal things


I recently interviewed with a 23andMe employee and specifically asked about the company's concerns regarding the public's perception of data privacy. The 23andMe employee was absolutely adamant that 23andMe would never sell customer data or allow anyone else access to it for commercial reasons. They said it went totally against company principles and that the company cares deeply about it.

And now this. I wonder if that employee was just not very attentive to what was actually going on, or perhaps the company leadership deceived the rest of the company? Or maybe this is just yet another example of a once-noble company inevitably succumbing to the allure of "growth at all costs".


You can always fake your name when you submit the DNA. Then this doesn't really matter.


they have to mail it to you


or can you? you have your credit card information surrendered.


Good luck getting an insurance when they got your DNA


- Business model? "... generating an estimated $475 million in revenue for the company, which has yet to turn a profit."

- Medical legitimacy? “Effectively what you have is a technology that is neither that helpful [nor] that harmful (...) Dr. Jonathan Berg, a clinical geneticist at the University of North Carolina at Chapel Hill."

- PR Signaling? “It was so unacceptable to her as a compassionate human,” says Ashley Dombkowski, (...) “She is undeterred by massive, worthwhile problems.”

- Personal Branding? Cue the Jobs-esque clad-in-black shots.

- 'Certainty' in the face of material uncertainty? "By October 2013, 23andMe was in talks with Target and Wojcicki was pushing hard to enter stores before the holidays (...) Emily Drabant Conley, recalls Wojcicki’s certainty when an exec thought it was impossible to meet the time line: “Anne was like, ‘This is a company that was founded on impossible.’” But in the end, impossible won. It would take three more years for 23andMe to get onto Target’s shelves. "

... so where have seen seen this before?

I'm glad that Forbes is taking a slightly more skeptical tone this time, at least calling out some of the issues and not taking PR at face value.

I actually hope they are successful, though I'll never participate due to obvious privacy concerns.


Hasnt anybody ever watched Gattaca ?




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: