"The partners plan to use 23andMe’s data to jointly discover drug targets."
They will claim all kinds of protections of course, but it is only a matter of time until genetic data starts being resold.
I think you're right to be concerned about sharing of genetic data, but conflating facebook / google data sharing practices with medical data sharing is not appropriate
If the concern is that 23and me may bring the google / FB approach to personal data to healthcare then i share your concern
Selling it to anyone who pays the price seems like a bad idea. Which is inevitably what will happen here the next time the company needs to "maximize shareholder value" by "utilizing its assets to their full revenue potential."
Next thing you know, I end up paying higher insurance rates because some distant relative I don't know is predisposed to some disease I've never heard of that is linked to a habit I don't have.
Again, because the insurance company is required to "maximize shareholder value."
> "maximize shareholder value"
> "utilizing its assets to their full revenue potential."
That's quite literally all that needs to be said here. How people fail to see the downside(s) to a large, multi-national, for-profit, shareholder-value-maximizing corporation, having access to (quite literally) their most sensitive information is simply beyond me. What do people expect GSK to say? "We plan to immediately redistribute this data to generate revenue - oh, and we do plan to do some research as well." GSK is not a charity. They do not run on merit, or "doing good things". They run on making - money. Any revenues from any type of wonder drugs developed via this 23&M partnership will be so far down the line (years, if they even come up with any winners). So in the mean time you can bet that they plan to get their ROI from this _investment_ via some type less than morally reputable activity.
It's shocking that so much genetic data is available, and we're barely scratching the surface, and using for fancy graphs and genealogy trees. We could be finding the cure for real diseases, and improving the lives of millions of people.
Yes, of course the pharma companies will milk the new drugs for the next couple of decades, until they lose the patent, yadda yadda, but at least we're making progress faster. Much better than the alternative - which is to wait decades for the discovery, AND wait more decades for the patents to expire...
Regarding privacy concerns, that's the least of my worries. The money available selling my genetic data (to, say, insurance companies), and the scrutiny and regulation they'd face (in many cases it'd be outright illegal - e.g., minors) is so massive that it'd be stupid to even try. There's a lot more money to be made from the exact same insurance companies by selling overpriced drugs to sick patients with health insurance.
Well there it is. You couldn't have an objective discussion about this, even if you wanted to. You (and your family) have already gone through the process. So of course you're extremely hopeful/optimistic that this works out with no problems ;)
23andMe is still held exclusively by private capital.
Nowhere in the reporting does it note GSK negotiated the right to resell 23andMe data. Furthermore, it specifically notes "[if an option is taken to extend the arrangement for a 5th year], GSK will become 23andMe’s exclusive drug target discovery collaborator."
Edit: See searine's thread below for a better discussion of why this is critically important to drug development https://news.ycombinator.com/item?id=17611775
I wouldn't be as worried about GSK reselling data to insurance companies. They already have access to sensitive health information for many many people and have had access to this data for decades. You literally cannot get a drug approved without collecting sensitive medical info on patients, because you can't tell if the drug works without collecting this data. Many drugs now require some genetic testing to be done to justify a prescription
There are many laws protecting against this type of data reselling and it also isn't really in a pharma companies business interest to help insurers. Healthcare companies tend to be very protective of and territorial with their data -- especially with enemies -- and pharma and payers aren't generally on the same side
I don't understand why there's an uproar over this but at the same time people are pushing for more medical data sharing among providers / payers etc for "population health". At the very least the absolutely certainty some people are expressing about this data being used maliciously seems unfounded. There is a massive massive difference between how data is handled in tech vs healthcare
For example: "They already have access to sensitive health information for many many people and have had access to this data for decades". In what country is this? Everywhere? You would be wrong about that, unless you mean a specific group of people that gave them this information voluntarily (in which case you made it sound a lot more exceptional than it is).
"Many drugs now require some genetic testing to be done to justify a prescription". How many? I've interacted with a lot of people who take (or have taken) prescriptions, and have met precisely zero that needed a genetic test. Of course this doesn't mean you are wrong, but maybe you have omitted critical data required for forming the opinion that you are correct. I already mentioned the number (and perhaps the type of prescription is also important). Also again, where in the world is this?
These aren't the only two claims I raised an eyebrow about, but they are a good start.
In the spirit of not wanting to seem that way myself:
* here is a link to all of the clinical studies currently being conducted by glaxosmithkline: [0]. There are 4,662 studies currently listed. This is across dozens of countries. Click around and look at the data they are collecting (look at the endpoints, inclusion / exclusion criteria, etc). This is all clinical study data, so patients have to volunteer for this, but you also have to opt in to sharing your data for 23andme. I also know of a few companies that collected lab data from patients and sold them to pharma companies so they could target them with ads; patients had to opt in to this and there were privacy protections; I'd imagine there are plenty of other examples like this.
Also, thousands of other companies have access to massive amounts of health data, genetic and otherwise: EMR companies, insurance companies, hospitals, etc. Many of these groups share information with each other all the time, including genetic information. How else could a lab share with a hospital a patients sequence data? Why is this GSK / 23andme deal particularly concerning? Is it because 23andme or GSK are viewed as bad actors?
* here is a list of FDA approved nucleic acid based tests: [1]. More and more drugs are being approved to treat patients with specific genotypes, see slide 316 [2]
[0] https://clinicaltrials.gov/ct2/results?cond=&term=glaxo&cntr...
[1] https://www.fda.gov/MedicalDevices/ProductsandMedicalProcedu...
[2] https://iabsverige.se/wp-content/uploads/InternetTrends2017R...
People underestimate how excruciating it is to collect large sample-sizes of genetic data.
The database 23andMe has is so stupidly valuable because 1. the sequencing is already done (and paid for) and 2. They can follow up with surveys electronically.
This kind of voluntary ancestry service will probably be the only way we will ever be able to collect the millions of samples needed for powered WGS GWAS analysis.
I think the 23 and me data is not sequencing data but genotype data. So it only looks at a certain type of mutation in a limited set of ~500k known mutations. I may be wrong so please correct me if so. So you won't find as many rare mutations in this data, or non-SNP mutations. Also I don't think they have robust clinical data for all subjects, it's just self reported. Again I may be wrong, I haven't done a 23 and me
This is a big deal for drug development. Each drug basically targets one protein. So you need a genetic marker that has a large effect size and is well correlated with a clinical phenotype. Not having clinical data is a big issue here. Also, 23 and mes database is not designed to find large effect size mutations -- their advantage is scale, and I think they only measure known mutations, so they are good at picking up common mutations with small effect sizes in common disease. The depression study they did is a good example of this application
But that type of study is low value for drug dev. You want large effect sizes. So if you have a big dataset, you want to find rare mutations with large effect sizes that are linked to extreme phenotypes, not common mutations with low effect size linked to common phenotypes. Basically finding more PCSK9 type mutations. Having only genotype data rather than sequencing data really hurts here
That's why I much prefer something like the Regeneron Genetic Center to 23 and me. They get robust clinical data, do while exome sequencing, and collect data from targeted populations where signalnis easier to discover.
In fact GSK was part of the RGC consortium but dropped out, dunno why. This may be their "rebound" from that
Again my assumptions about 23andme may be off bc I haven't used their product
>> The Regeneron Genetics Center (RGC) has built one of the world’s most comprehensive genetics databases, pairing the sequenced exomes and de-identified electronic health records of more than 300,000 people so far. https://www.regeneron.com/genetics-center
https://en.m.wikipedia.org/wiki/23andMe - users 5 million
My understanding is that you are saying 23andme's 5 million records are less useful for drug discovery than 300,000 records with
1. more detailed / reliable medical records 2. they do the wrong type of genetic analysis (which i don't fully follow)
Could you expand more? Can they re-use their genetic material if it was all stuck in a fridge and do the "better" sequencing?
Specifically, that the FDA prohibited them from proactivitely notifying users about potentially dangerous mutations. Looks like they've gotten approval for a few breast cancer risk markers now, but that's well short of what they could do.
If you're notifying users about a drug discovery program... the FDA can't gripe.
My argument: If you are exploring genomic datasets to find new potential drug targets, then what you really want as the output are single genes that are very strongly associated with dramatic phenotypes. Drugs generally only hit one molecule, so you want a monogenic vs polygenic signal, and drugs can't hit every molecule of a given type in the body, so you want a large effect size (if you only hit 10% of the target you still want a meaningful clinical effect)
If you don't have clinical data, your genetic data is Not super valuable for target discovery. You can't correlate genetics to clinical outcomes. The more clinical data you have, the more valuable your dataset, because you can uncover more gene-phenotype correlations. So you need high quality medical records, lab data, etc. if you just have self reported data on a few diseases, you'd miss all sorts of potentially useful signals
The type of genetic analysis is also super important. 23 and me does "genotyping": they have a chip with like 500k-1M molecular probes. Each probe is designed to detect a specific "SNP" mutation, i.e. A mutation where one DNA "letter" is different. So this doesn't pick up other non-SNP mutations but that's not as important. What's more important is that there are like 10M+ (edit: prev said 360M) known SNPs and prob a lot more unknown ones. So with 23andme you are only exploring a small part of the genome
And this part of the genome is fairly well explored. While next gen sequencing is a newer tech, gene chips (what 23andme uses) have been around longer. Most common mutations have been studied. And 23andme is just studying those common mutations but at a larger scale
However larger scale isn't necessarily that great for target discovery. With larger database you can pick up 1) more relationships with small effect size and 2) more rare relationships with large effect size. Except 23andme is using a gene chip that detects mostly common mutations, and bc they have limited clinical data they will mostly have common clinical condistion in their dataset. So you can really just pick up a lot of common mutations with significant but not meaningful relationship w outcomes
If you use exome sequencing like RGC, you get much richer coverage of the protein coding genome than 23andme. So you can pick up rare mutations. And you have more clinical data (arguably having more clinical data per genome is more valuable than having more genomes), so you can pick up more gene-phenotype relationships. You need to scale your sample set so you can detect rare mutations, but do you need 5M people? The more the better but RGC has already yielded some promising targets w it's smaller dataset
RGC is also smart and targeted about the kind of patients they recruit. So there is less noise and more signal, so you don't need as many patients. For example they look at fairly genetically homogenous "founder populations" that have less background genetic variability. Like the Icelandic population -- PCSK9 was discerned by analyzing this pop
PCSK9 gene is a classic example: mutations in this gene are very strongly associated with extreme levels of LDL cholesterol. And the relationship works for both gain of function mutations and loss of function mutations, and the causality can be validated experimentally.
The effect is dramatic: patients with loss of function mutations in PSCK9 have like 10% of the normal level of LDL cholesterol
And it's a monogenic trait: you can get these extreme LDL levels just by modifying PCSK9
So this is a great target assuming you can design a molecule to "block" it (you can). You can create an antibody that can basically have the same effect as the mutation (keeping PCSK9 from doing its job) but on a smaller and less durable scale.
The drug worked at lowering LDL cholesterol. It has had mixed commercial success for a variety of reasons unrelated to its effectiveness of lowering cholesterol
So this was really the first drug discovered based on large scale genomics. Regeneron developed one of the two PCSK9 inhibitors on the market. They purpose-built RGC to find more of these
I presume that GSK thinks they can do more detailed follow up medical questionnaires on the 5m existing users than just any random sample.
But will the genetic material still be viable for retesting with the more advanced chips?
Also it would cost maybe ~$500 / person to do whole exome sequencing, so doing this at scale is quite costly. Thats another reason you need to be targeted when recruiting populations for sequencing, it is so expensive. also RGC spent a lot of money on a high throughout, heavily automated sequencing center, so they can sequence more cost effectively than most. Even the UK Biobank genomic project is using RGC to sequence their patients
[#] insert your high quality long form news source here.
That's what the goal of https://allofus.nih.gov/ is. To collect data from 1 million people in the US.
Seems dangerous.
While it's also possible that advertisers could use more sensitive information, such as a predisposition to gambling/obesity/alcoholism, the signal from DNA that an advertisement is likely to be effective is still probably weaker than a person's actual habits. A very unethical company could use DNA to know how to target to make "the first hit free", but I'd like to think that such egregious behavior would result in significant public backlash and possibly the introduction of regulations
Edit: As people bring up the insurance risk, yes I did consider that and mention it in my post. I am personally not concerned about it and think the advantages outweigh the risk. Besides, I think insurances companies are going to have all the data soon anyway. If every time I get a blood test, they have a chance to capture my DNA, how can I stop them?
There are laws against age, race, and other discrimination in employment. But it still happens, and in some industries in widespread.
Greed finds a way. Your point is moot.
While adverse selection is undeniably a tricky problem, the underlying raison d'etre is pooled risk, to help individuals and families mitigate catastrophic losses. Maybe you're all right with the prospect of future analyses condemning to uninsurability folks with rare combinations of alleles that turn out to be strongly deleterious. To me, this seems like breaking the regulated semi-statis between insurers and insureds, and like breaking the social contract more generally.
The whole system is going to have to adapt to more information on individuals being available.
This might or might not be an issue now (regardless of the place of residence), but laws can always be changed. And societies can always develop to be more totalitarist, even up to extremes, and it is not like the collected data just vanishes when things start to slide. (Case in point: IBM's Hollerith machines and population records)
I was very curious about 23andme and really wanted to try it. After some due diligence I concluded it is not worth it because of privacy and other implications going forward. It is not what they do today, since they don't hide the obvious sharing and datamining, so if one joins them it is through informed consent. My concern was rather what kind of systems and policies they enable eventually in the future, after the information gets sold/shared enough times to anyone who can pay enough. At this point there is no control anymore in how (and by whom) the data gets used.
Cue adverse selection. They might not be allowed to collect genetic data, but nothing stops patients from doing their own sequencing, and shop for healthcare plans depending on their predispositions, thus drastically affecting healthcare profitability.
This is not a pervasive problem yet because the tests are pretty crappy and very few people bother, but on the long run, I bet your "level playing field" hypothesis will be debated.
How about displaying recruitment ads to people with lactose tolerance gene, effectively excluding most of worlds non-white population?
> It shall be an unlawful employment practice for an employer-- (1) to fail or refuse to hire, or to discharge, any employee, or otherwise to discriminate against any employee with respect to the compensation, terms, conditions, or privileges of employment of the employee, because of genetic information with respect to the employee;
No job advertising protection there
NJ (LAD):
> It shall be an unlawful employment practice, or, as the case may be, an unlawful discrimination: a.For an employer, because of [..], genetic information, [..] to refuse to hire or employ or to bar or to discharge or require to retire, unless justified by lawful considerations other than age, from employment such individual or to discriminate against such individual in compensation or in terms, conditions or privileges of employment;
Nope, nothing there either about recruitment ads..
I think the pros far outweigh the cons.
Sure, 23andme and GSK and others will get richer off this but so what—they are doing something good.
I think few people would argue against the benefits of genetic testing, but just because it's a great thing doesn't mean that being complacent around adjacent shady business transactions is the right approach.
You can be for it, but also for privacy, protection against insurance discrimination, etc. at the same time.
I’m dead in the long run regardless, and would rather find every way I can be impactful with the time I have left, even in these small ways.
10 million tests vs 10 million person database means 1 in 10 trillion false positives happen 10 times.
Generally at a 7cM overlap you have 50:50 chance whether you're related to that person or not. Anything over say 150cM you're virtually certain to be related (ignoring endogamous populations). At 3,400cM (full-match) it's not even worth considering randomness.
Paternity tests involve random ~50% of your dna. Sperm samples are generally contaminated and thus less clear cut. But, people’s relatives have been used in the past for these tests dropping the bar even further.
However, being related genetically is not enough. Some people have twin siblings they don’t know about pointing out being related is not nessisarily mean you know anything about the other person allowing for false positives even at 3,400cM.
Unlikely sure, but harm comes in many forms. Saying I can’t think of anything is a long way from saying it’s safe.
If you've got a identical twin then you're right that their child will be genetically indistinguishable from your child. But that's down to the nature of genetics rather than anything to do with the reliability of testing or database size.
If a test failed because _ then it failed independent of why it failed. If your long lost twin causes you to be involved with a paternity suit or homicide investigation then that's both a false positive and a significant problem.
Suppose a male with red hair, green eyes, and AB+ blood actually has those traits they are just not enough to unlikely identify someone. Adding more genetic traits on it's own is not enough for example you have identical twin or even triplet separated at birth. The core issue is DNA marks are not independent though people often assume they are.
Sure, at some point in the future it might get fixed or always fail, but until then you need to deal with today not some mythical perfect future.
For more hypothetical societal risks, I’ll cite the movie Gattica where DNA credentialism has created an underclass of people who can’t afford genetic manipulation at birth of their children. If your DNA is found wanting in that society it could be used against you in job applications, rental applications, etc.
Currently (IANAL) but I don’t think this would be legal, but once your information is out there, it’s rather hard to bottle back up.
Currently, it's explicitly illegal under US federal law, but there is also a full-frontal assault going on against consumer protections in the health insurance market in the federal government, so absent a significant change in government direction, I wouldn't bet on it staying that way.
Insurance only works as a system for society if the risk is spread among many parties, but if the insurance companies know the risk ahead of time it ceases to subsidize the unfortunate, who are left out in the cold (taxpayer funded programs and hospital ERS)
I’d make an exception for science if and only if it’s full deanonymized and controlled by a nonprofit third party entity.
People who have serious disabilities should be supported by taxpayer funded programs, in my opinion.
Also I hope you never have my DNA. I can’t think of a reason why you should, and for one possible exercise on why you shouldn’t, please watch Gattaca.
As for your DNA: Do you take the precautions of the protagonist in Gattaca? How do you avoid leaving DNA samples in public places? How long before drones/bots/Roombas are hoovering this up at the behest of some VC?
Re DNA precautions: We just need to make hovering up DNA illegal, or selling insurance, or selling anything based on the genetic information gathered from people; make it all illegal, so I don’t have to worry about it in the same way I don’t have to worry about insurance companies breaking into my house to spy on me.
What happens when someone gets their genetic information released in a country with public healthcare, and then moves to a country with private healthcare?
Why is that bad?
US insurance companies use treatment guidelines issued by organizations like NCCN (oncology). If they stray too far, they get dinged for not meeting "medical necessity" requirements.
If anything, you're more likely to get an experimental cancer drug in the US than Europe since US insurance companies don't like bad press.
“Based on your genetic profile you’re 27% more likely to develop ________, talk to your doctor about how GSK’s _________ can help”
Having the same information available to your doctor (or an AI, ha!) that could parse it objectively would make more sense than a single pharma company. Does anyone really think that GSK is going to suggest you take Viagra (Pfizer) over Cialis (GSK product)?
“Based on your weight you’re 27% more likely to develop heart disease or cancer, talk to your doctor about how diet and exercise can help”
Which I personally think is a very good thing, but a lot of other people see as "fat shaming" and a gross violation of their human rights.
People bang on about early intervention, and it's useful for some stuff, but for a lot of stuff there's no evidence of a reduction in all cause mortality.
For example, we currently do not have a diagnostic test that can tell the difference between a virus or bacteria so physicians prescribe antibiotics just in case. If we did then we could reduce the use of antibiotics. Genomic sequencing won't help this case specifically, but sequencing at the proteomic or metabolic level may.
1. A lot of people come to a doctor with a specific prescription medication in mind based on ads alone and many doctors happily oblige and possibly even receive kickbacks from pharma.
How often is medication prescribed for a problem that could have been left alone? I suspect A LOT, based on how over-prescribed addictive and dangerous opiate painkillers are in USA.
We'll now be corralling those people not just on conditions they have, but also ones there is some research suggesting they are a higher risk to get. Maybe they'll even start taking something preemptively, because they have a 200% higher than average chance of developing a condition that affects 1 in 1M...
2. Once your genetic traits data gets to the adtech it's a matter of time before you'll see online gambling ads targeted based on research like this https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3638385/ and similar things
3. Negative targeting could also become a thing:
- how about you don't show health insurance ads based on genetic markers that can result in high coverage costs? or limiting your recruitment ads to potential employees that aren't likely to have or develop diabetes?
- just checking for lactose tolerance gene would exclude most of worlds non-white population - would that be OK? that same check could be used for medication ad targeting - wouldn't even have to extended dataset
I'm fairly sure I could get you a hundred examples how it could go terribly wrong if I had a couple hours.
The general answer is that medicine is supposed to help you. But advertisers aren't paid to help you. Theyre paid to get you to buy things. And all the trickery and marketing tactics aren't conducive to getting you the correct treatment.
email I received from 23andme today. "Our top priority is you, the customer, and empowering you with the options to participate in research. As always, you choose whether or not to participate in research. You can choose to opt-in or opt-out at any time."
If you're not a customer, might not want to throw around assumptions.
Give away or sell tests at a loss to vulnerable cancer patients in search of hope, hoard the data, and sell it to the highest bidder.
Or hacked.
Is the hacker going to call me up and suggest I start taking high blood pressure medication?
Also since we're talking criminals, blackmail: eg tell your employer about your predisposition to violent mental illness or reveal your kids aren't actually your spouse's.
Or stolen. Or both.
Is there a way for them to use the data "to jointly discover drug targets" in aggregate without names/identifiers?