> Roelof Botha, a 23andMe board member and partner at Sequoia Capital, said the company’s big-spending strategy made sense when money was cheap. Now that it isn’t, “we’ve had to trim and focus on a smaller number of projects.”
> Sequoia, which invested $145 million in 23andMe, still holds all its shares, he said. Today they are worth $18 million.
Interesting that they weren't forced to liquidate/distribute the shares to their LPs
I don't know what specific deals they have with LPs, but why would it make a priori sense to force the fund to liquidate some specific shares and distribute the proceeds?
I would imagine that any fund is allowed to make losses on specific investments, least of all a hyper-aggressive VC fund which will create losses on many (most) positions, by design. And then once losses happen they should be treated for what they are - just another position where decisions should be made based on the outlook, and not based on the history.
in 2021 Sequoia switched their strategy to hold on to stock of their portfolio companies for few more years after IPO.
"Sequoia is abandoning the 10-year venture fund, in which limited partners, the outside investors that contribute to the fund, expect to get paid back over a decade. The firm said it’s establishing a single fund, the Sequoia Fund, that will raise money from LPs and then funnel that capital down to a series of smaller funds that invest by stage.
Proceeds from those funds will feed back into the Sequoia Fund. With no time horizon, Sequoia can hold public stock for longer stretches, rather than distributing those shares to LPs. Investors who want liquidity can pull money out instead of waiting for distributions."
but more of an object lesson that as the needs of the company change, the board/your co-founders will constantly re-evaluate whether you are still adding value.
No company that had to go public via SPAC is going to bother with all that. The whole point is to pump and dump. This company will be folded into some drug company for its "IP".
Seems like 23andMe is two businesses: consumer and a B2B data business.
The consumer side is clearly struggling because of the problems mentioned in the article (they only need one test in their life, public perception is bad because their security has had breaches). So this needs a pivot where you can change the public's perception from a one-time test to continuous health monitoring through blood markers or something similar, expand to tests other than genetic and make it a repeatable, accurate test that gives you more information (and obviously stop leaking people's data).
But why not focus on the B2B side? Sell access to their databases. I'm sure computational biology and/or pharma companies need this information. It makes sense to do vertical integration by manufacturing your own drugs, but not for a cash-strapped business that has little incoming revenue to sustain further development. Presumably they are selling their genetic data, but I don't get why it's not giving them a revenue stream. Let GSK manufacture the drugs using your genetic info, with a profit share for any drugs made this way. They mention a collaboration with GSK in the article, but why was this stopped?
> needs a pivot where you can change the public's perception from a one-time test to continuous health monitoring through blood markers or something similar, expand to tests other than genetic and make it a repeatable, accurate test that gives you more information (and obviously stop leaking people's data).
"Stop leaking data"? Sounds like a step forward.
> But why not focus on the B2B side? Sell access to their databases.
So, it's actually: try to sell your customers' data instead of leaking it? :O
Sorry if that sounds snarky, but are we sure there are enough customers who want to pay to give their data to a company so that company can immediately sell their data on to other companies?
That's the conflict at the heart of the business. But do the public care enough about that? Obviously a HN audience does, but people use Google and Facebook products every day with the awareness that all that data is sold directly to advertisers. With the right messaging ("yes, we sell your data, but it's to drug companies to help make drugs that can cure your illnesses") it's possible that the conflict isnt too much of an issue.
Drug companies pay pretty well for clinical trials, so why would anyone pay 23andme for their own data, so that 23andme can turn around and sell it to drug companies?
> But why not focus on the B2B side? Sell access to their databases. I'm sure computational biology and/or pharma companies need this information.
Are you sure?
I mean, presumably there are different types of DNA testing. Doesn't 23andme run basically the cheapest test they can get away with? A user can't tell if the test measured 16 bytes or 1.6 gigabytes of genetic information, and if I was trying to launch a consumer DNA geneology service, I'd want to get network effects, so I'd want a test that was very easily affordable.
Who says their records are thorough enough to be valuable to drug companies?
You make a good point. It's possible it's not useful to them and that's why they can't generate enough revenue from it. I assume the sales pitch to pharma is that it's a wide database, rather than deep.
> Doesn't 23andme run basically the cheapest test they can get away with?
No. That's part of why they're test is so expensive and their financials are so lackluster. They are apparently using a customized version of the lllumina's Global Screening Array according to their website and several other sources that show up in search results. That's a legit research quality genotyping platform from a world leading laboratory in the genetics space. This post from 2020 has a decent high level overview about it in the context of 23&me [0] though it might be slightly outdated by now and I've never heard of the company (xcode?) that wrote it (nor did I bother to look at what their product is.)
> If I was trying to launch a consumer DNA geneology service, I'd want to get network effects, so I'd want a test that was very easily affordable.
That's a pretty neive perspective. It ignores the value propositions of 23&me's product, economies of scale in the direct to consumer genetic testing space which were in large part enabled by 23&me's success, and all of the thorny bits related to questions about accuracy when presenting results. Not to mention that 23&me certainly capitalized on the network effect (which is being criticized a fair amount in this thread.)
> Who says their records are thorough enough to be valuable to drug companies?
> Presumably they are selling their genetic data, but I don't get why it's not giving them a revenue stream. Let GSK manufacture the drugs using your genetic info, with a profit share for any drugs made this way. They mention a collaboration with GSK in the article, but why was this stopped?
In 2018 GSK made a $300M equity investment in 23andMe as part of a 4 year collaboration (with the option to extend for a fifth year) under which GSK had exclusive access to their data for use in drug target discovery programs, but [0]:
> All activities within the collaboration will initially be co-funded (50%/50%), with either company having certain rights to reduce its funding share for any collaboration programme.
So it seems they not only lost out on 5 years of developing their B2B business, but committed to covering a portion of the R&D costs over that period as well. There were terms about profit sharing on new developments, so it was a bet.
It doesn't sound like it worked out quite as well as either sided hoped though because in October 2023 (after the 5 year agreement) they entered into another agreement but this time [1]:
> Under an amendment to their Collaboration Agreement, 23andMe will receive a $20 million upfront payment for a one year, non-exclusive data license.
> [...]
> for a 12-month period, and [23andMe will] offer its research services for analyses of the data over that same period. Any new drug discovery programs that GSK chooses to initiate during the agreement will be owned and advanced solely by GSK.
The real value they tried to build was in pharma, which is much more lucrative than either of those. Their data is somewhat interesting but limited by the lack of rich phenotypic/clincial information paired with the genetics.
> She’s sticking to her goal to transform 23andMe from a supplier of basic ancestry and health data into a comprehensive healthcare company that develops drugs, offers medical care and sells subscription health reports.
The company hasn't the funds (or time) to develop drugs and I'm not sure why any investors would invest in them rather than a company with a track record of doing so or a potential drug.
Selling their customers data to other companies (or partnering with) that can create drugs from it would seem the logical choice, along with "subscription health reports" and anything tacked onto the DNA/blood profiling/testing product to keep the data coming in.
I would have thought they were pretty well positioned to find new targets for currently untreatable diseases. But maybe they don't quite have the capacity needed to target specific diseases and/or data not granular and clean enough, or not enough capital/runway suitable for that kind of stuff. If they had the sequences for people with diseases such as alzeimers/Parkinson, they could just look for common mutations and see what specifically it affects
It is not like they are doing real genomics. It’s devilishly hard to find genes responsible for conditions like schizophrenia, heart disease, asthma, diabetes, etc. The last thing you need is mountains of low quality data.
Attempting to discover drugs is a proven way to make a small fortune by starting with a large fortune. Everything north of the Charles River in Boston would vanish if people stopped believing though.
I don't think 23andme's strongest asset point would be in direct drug discovery, but rather in helping target sub-populations for clinical trials. The SNP data that 23andme has is relatively low quality compared to proper sequencing, but (combined with their survey data) is probably at least as good, or better information available for typical clinical trial planning or screening.
> I don't think 23andme's strongest asset point would be in direct drug discovery, but rather in helping target sub-populations for clinical trials
Was there an issue with targeting sub-populations for clinical trials beforehand?
At an ELI5 level, if you're hoping your drug candidate will help cure disease X, you sign up patients with disease X to join your clinical trial. That's not the hard part!
The two (related) cases where you'd perhaps want genetic information is:
a) If you suspect there is a significant pharmacogenetics component to what you are studying (or related to disease progression).
b) You're working on something preventative.
From a previous job I worked at (we made PCR tests), there was interest on screening for APOE genotypes to enrich an Alzheimer's drug trial - drug maker believed that APOE genotype would have a significant impact on drug performance.
But you're absolutely right that you can often (usually?) do 'enough' enriching without genetic information.
My understanding is, ideally your clinical trial manages to capture or balance out the different potential factors in genetic variation.
An example of this is when they tested the covid vaccine, they wanted to make sure they had enough participation from African American and other ethnicities since these were usually under represented in clinical trials relative to the population, and there are sometimes subtle variations in the way peoples body respond to drugs depending on their race.
> I would have thought they were pretty well positioned to find new targets for currently untreatable diseases
To find new targets yes but to develop a drug to market (and revenue/profit) takes 10-15 years. Even a drug to a state that gets approved for human testing seems beyond what they could do themselves without so much funding it would wipe out the existing shareholders. Selling to/Partnering with other drug companies with targets they've discovered/IP from DNA etc would seem more realistic.
I would say target discovery is definitely no less important than drug discovery though. Biology is so complex, for a lot of diseases we aren't even sure the targets and pathways are valid. But then again, as per another commenter, I guess it's much harder than it seems, and probably not as easy to monetize as a drug
Trust me, it's not. I've seen start-ups worth 100x than a smaller successful business due to speculation, despite the smaller business employing more people and actually turning profit for many years.
There’s plenty of successful businesses chugging along growing at about the rate of inflation. It’s really hard to grow a business at 10 or 500 times that rate.
Why would it be surprising to value the latter kind of business higher than the former?
Growing at the rate of inflation only matters if your market is capped out. Like, say Soda makers.
If you either create a brand new market, or are taking on a drastically undeserved market... you absolutely need to be growing faster than inflation. Or your brand new company would be worth $1 the first year, $1.03 the second year..
Yeah I was trying to convey there are successful businesses that are broadly maintaining and there are rocketships (and everything in between) and we should expect them to be valued differently and have different risk profiles. So the observation (100x! That’s dumb!) is not rational.
VC logic is that 5% of the companies they invest in returns 100x their investment in those. So even if 95% of the companies never manages to break even they are still profitable.
VC logic is that a company does not need to be profitable for a VC’s investment to be profitable. Simply getting the company to an expensive IPO is all that is required for VCs to make money. And that is possible even if the company has never made a dime.
This is the case for all investment, not just VC. Even cash, gold, real estate etc works on the same principle. It is valuable because other people find it valuable, not because the thing will give investor the profit in itself, just the valuation are expected to grow.
> There’s plenty of successful businesses chugging along growing at about the rate of inflation.
Even outpacing inflation by 1.2 - 2 times. It's not insane growth, but it is consistent. Better yet is that they tend to have a decent number of assets too.
> It’s really hard to grow a business at 10 or 500 times that rate.
How often is 10 to 500 times growth seen in reality? I think most start-ups fail before the second year.
> Why would it be surprising to value the latter kind of business higher than the former?
Highly stable and proven growth as opposed to highly unstable and unreliable growth.
Remember that startup "valuations" are not what they're "worth" - it's simply how small of a segment they have been able to convince the last round to take. The valuations have a bunch of unknowns that can wildly swing it.
The small business is successful and so the calculus as to what it is worth is much more explicit.
The extreme example of this is commercial real estate with a long-term tenant (think: the person who owns the land a post office is sitting on). Everything is known (lease length, amounts, current interest rates, etc, etc) that the value/sale price is a simple formula.
Richard Branson, Oprah, SNL, billionnaire neighbour investor, Google founder, celebrity parties, branded Barbie dolls and SPAC IPOs can take you a long way .
Oddly, 23andme adopted an extremely cautious, risk-averse strategy with respect to providing "health information". The result of this is that many other services became a more interesting/entertaining source of personalized health information based on genetic data (often 23andme raw data extracted and fed into the other system).
I get new updated health information from 23andme perhaps 2 times per year. It's mind boggling to imagine that with all the research being done that is all that makes the cut as interesting and trsutworthy enough for 23andme to tell me about it.
In other words, a majorly failed content strategy based on fear of being shut down for providing insufficiently vetted health info.
Meanwhile, someone created multiple accounts and the ancestry results didn't match even when collected from the same person.
No, but I would prefer greater transparency about the accuracy of the analysis and also greater transparency about the reliability (or unreliability) of any health recommendations based on research.
I'm not looking for a single source of authority and I will not be disappointed if an updated "chip" or study reverses a previous recommendation/insight.
23andme had a blog post about the MTHFR gene and why they don't include it. I'm not there to read the blog I want to know interesting things about my own genetics, even if what is interesting is that they don't think it matters! This is just bad content strategy even if it's justifiable from a "healthcare" perspective.
Why so much worry about a customer getting the wrong impression from one of the health reports? Seems a bit absurd. How much money did it cost to put in place all the warnings about the potentially upsetting impact of various mutations.
I wish that 23andme had focused on being an entertaining way for people to discover more interesting stuff about themselves, rather than a rarefied place where only the most well-vetted research is acknowledged to exist.
This story makes 23andme sound like a non-fraud version of Theranos, with some of the same issues around business viability, dodging the FDA, relentless PR and funding from rich friends.
If a healthcare startup get shout-outs from Oprah, SNL and branded Barbie dolls, is that a bad sign?
If Theranos devices had been able to do what they claimed the business model was clear and likely very profitable.
I don’t understand how anyone ever thought 23andMe had a viable business model selling a test that only needed to be done once(and apparently selling it at a loss to boot)
The "if" was the entire scam though. They hired many smart people, and every one of them said that the proposed design was physically impossible to build.
There's a reason that these complex undertakings are made by boring conglomerate companies that have decades of experience in the industry, produce many different products in that sector, and whose CEOS don't try to get on the covers of magazines.
Precisely. Yes, breakthroughs get made by smaller companies and "outsiders" every so often, but the thing is, to these testing companies - there's billions in profit and improved logistics by figuring out "nano" collection. It wasn't that they weren't interested. They've tried, for decades. But in the end, biology and physics are winning out.
It's not clear to me that this was in any way a money-making business. The test only needs to be taken once. Selling this data to 3rd parties by way of a sneaky EULA may be part of SV culture, but it's probably not something that would get you repeat business in healthcare.
It's weird to me because the test cost was always more than enough to make it profitable. This is the kind of product where you can actually make a profit on every sale (funny I have to make that remark but it's the 2020s) ... and then ... they didn't somehow? :/
They cut the cost of the test from $399 to $99, so maybe their margins were not as good. Also, it would be a red flag to investors if there weren't any tacked-on "services" on top of the main offeirng to keep squeezing customers after the initial sale.
That is a pretty awful click-bait title. Presumably the "$0" is in reference to their sub $1 share price?
The fact that they put everything behind a paywall _and_ slap on click-bait titles is confusing to me. I can't imagine paying for access to that type of writing.
This is a classic combination of cargo-cult matching with ZIRP regime.
23andMe never had a business plan, but it didn't matter because their investors figured they had the connections to make it work (look at the explanation of how they booted out the other founder). Their Libertarian/lazy attitude towards regulation is just an example of how couldn't be bothered to even do the necessary homework. Their attitude towards their data breach is another.
Makes me wonder how good their science is. I don't know anyone on that side of their business, but comments from friends who used to work their about how their parts worked are not encouraging.
> To create a recurring revenue stream from the tests, Wojcicki has pivoted to subscriptions. As media companies launched streaming “+” channels, Wojcicki rolled out 23andMe+, offering personalized health reports, lifestyle advice and unspecified “new reports and features as discoveries are made” for an initial $229, with annual renewals of $69.
I was a heavy believer of 23andMe until this point. I answered all of the available research questions, which was a thing that took absolutely hours and was filled with semi-invasive medical questions. I did this under the premise that I would hopefully be helping research and I felt really rewarded having completed all of them. Then, they dropped the + bombshell and I felt really rugpulled. I paid them for genotyping on their v4 and v5 platforms -- so I paid twice, I referred friends, I bought people kits, I helped research...and now I was being asked to pay a subscription for what I was promised to begin with? Eesh.
I think the fundamental issue is that when the FDA stepped in and told 23andMe to cut back on their reporting that they really hit a roadblock. Promethease always gave better results than 23andMe, but when that happened it commoditized genotyping fully. Just download your data from any provider and use a third-party tool and you're set.
Fast forward to 2024 and they're stuck with a platform that's semi-limited and they haven't delivered on any of the research deliverables that many people wanted in the first place. The idea of getting new genetic reports monthly was appealing and simply never materialized. It's no surprise people aren't hot on this as a business -- but what is surprising is that they were completely passed by startups like Nebula Genomics in offering whole-genome sequencing and competitive data access. I think the stored data they have is their only advantage, but they don't seem to know how to leverage it.
> think the stored data they have is their only advantage, but they don't seem to know how to leverage it.
I really wonder what happens to that data if they get acquired or shut down. The possible new owner of the data might have a completely different business case in mind. This scenario is something you usually don't consider when you give your data to a company that you trust at the moment.
It's illegal in some US states like California. The California Genetic Information Privacy Act prohibits these companies from disclosing a consumer’s genetic data to any entity that is responsible for administering or making decisions regarding health insurance, life insurance, long-term care insurance, disability insurance.
Promethease is why i initially did 23andme, and it was beyond worth it for me personally.
But for the same reason I'm just a one time customer, all I needed was for 23andme to give me a very expensive hospital test at a 95% discount, and I didn't need them anumore
As someone who paid for 23andme+, it was pretty much a scam anyway, I got maybe 4-5 more reports over the year, all of which were for random popular ailments like "Anxiety" where the link between your genes and the disease(?) are pretty dubious in the literature and the takeaway was "you are 5% more likely to experience anxiety".
> where the link between your genes and the disease(?) are pretty dubious in the literature
This is inescapable; genetics usually can't show causation, for instance because you can't do an experiment where you change someone's genes.
Geneticists seem to deal with this by using statistics like GWAS that are obviously just correlation, adding a sentence that correlation doesn't show causation, and then just proceeding and pretending like it does.
Sure, perhaps that was a bit too succint. What I mean is that, as a customer of 23andme, I'd much rather know about some known genetic disease that I may or may not have than about wide-ranging things that affect everyone like stress and anxiety. It's just not as serious.
23andme went with the "only give very limited information" which I respect but is ultimately pretty much useless compared to just running your data against dbSNP and telling me that I have this or that marker for this or that disease backed by this or that paper.
Only showing correlation is one thing, but if the correlations themselves are barely noticable then that's a big problem for making a useful report.
Though if you find whatever gene is most correlated with something, what are the options for it not to be causation? If the chance of causation is high enough, it makes sense to proceed as if the risk is real.
When I said correlation isn't causation I meant it. Neither high nor low correlation is evidence for causation.
There's a more advanced form of being bad at this where you think you can show causation by controlling for everything in the environment. This is also wrong; it produces something called collider bias.
> Though if you find whatever gene is most correlated with something, what are the options for it not to be causation?
1. It's a coincidence and it's never causal.
Imagine an OSS project releases a bugfix and you diff the old and new versions. (This is basically GWAS.)
The bugfix part of the diff caused it to be fixed. The updates to the copyright dates or changelogs didn't.
2. It's causal, but the causal chain involves a specific environmental factor, and we should change that instead.
For instance, you can say every human has a genetic disease that prevents them from producing their own vitamin C, which most other mammals can do. But instead of calling scurvy a genetic disease we just eat fruits and vegetables.
Half the point of analyzing statistics is to filter out coincidences, and that applies to correlations too. If something is a coincidence, it won't hold up as a proper correlation under reasonable amounts of analysis. So when the premise is we're starting with correlations, I think it's alright to assume they're mostly not coincidences.
> The bugfix part of the diff caused it to be fixed. The updates to the copyright dates or changelogs didn't.
In that case the bugfix is the "gene most correlated", isn't it? Give it a few generations to randomly spread, and the signal will be far stronger on the bugfix gene than on the copyright gene. (And if it hasn't been spreading for generations then you won't have enough samples to find either gene.)
> 2. It's causal, but the causal chain involves a specific environmental factor, and we should change that instead.
> For instance, you can say every human has a genetic disease that prevents them from producing their own vitamin C, which most other mammals can do. But instead of calling scurvy a genetic disease we just eat fruits and vegetables.
If you're testing just humans, you'll get a 0% correlation because everyone has that gene.
If you're testing across mammals, then "WARNING: Prone to scurvy". Which is a completely correct and causal result about a genetic problem, with easily accessible treatments.
Please have some awareness of how little thought you are putting into dismissing a field where thousands of very smart people are working on solving the problem you describe. In fact, you can correct for the exact problem you describe with sibling studies where people have the same environment but different genes.
Have maybe a small ounce of humility in this respect.
> Please have some awareness of how little thought you are putting into dismissing a field where thousands of very smart people are working on solving the problem you describe.
Thousands of very smart people do a lot of dumb things. People still work on string theory. The people working on proper causal inference are also smart, work hard, and have Nobel Prizes.
On the other hand, the people doing genetics who aren't careful about it produced the 23AndMe report which says I have a "17% chance of having green eyes". Against what counterfactual?
> In fact, you can correct for the exact problem you describe with sibling studies where people have the same environment but different genes.
"Correcting" is a wrong way to think about it. Generally speaking, overcorrection is worse than undercorrection because of collider bias. You need to choose a study design that's correct in the first place.
You're describing a natural experiment, which is better than a GWAS of self-selected 23AndMe customers, but does have problems (silly one: selection bias because the sample only includes people with siblings) and more importantly is only guaranteed in the study environment (eg people who live in the UK in 2008) but gets reported without identifying what that environment is.
> > In fact, you can correct for the exact problem you describe with sibling studies where people have the same environment but different genes.
> You need to choose a study design that's correct in the first place.
That's precisely what a sibling study is: a study design that's correct in the first place, as far as the problem we're discussing goes.
> You're describing a natural experiment
That is correct: a sibling study, which corrects for the problem we're discussing, often contains aspects of a natural experiment, namely that the siblings tend to naturally be siblings, rather than raised together purely for the purpose of a study :)
> As an added security measure, we have temporarily disabled the ability to download your raw genetic data. We hope to re-enable this ability soon, and we appreciate your patience.
Apparently 23andMe doesn't let you download the data anymore, I just tried. I wanted to give Promethease a go, seems interesting.
"We have temporarily disabled raw data download as an additional precaution to protect your privacy. We don't currently have a timeline for when this feature will become available but will keep customers informed of any changes."
It is an embarrassment that they bought Lemonaid health, I don’t know any other way to put it.
(1) Is that being associated with a penis pill pusher is brand destruction. Any normal person would get this right away without trying but executives just can’t see what their actions look like to their customers.
(2) It’s a form of achievement laundering. Lemonaid health failed, pure and simple. A circle jerk acquisition is alchemy that turns a failed business into an apparent success, now the founders, management and people who invested in it will get accolades for being successful when really the Silicon Valley Mafia stole from public investors to maintain its image.
It ought to be a lot harder for businesses to get acquired. Maybe we need a vote by the customers, maybe when Google buys a company they should be required to spend a 1/4 of the money on ads warning people that the business could get shut down, etc.
Their kit is 99 dollars, but how expensive is the actual DNA testing nowadays ? It's my understanding from this [0] video that it is approximately 1000 dollars for the whole genome, but 23&Me sequence only a tiny portion of the genome (above video mentions 500 000 bases out of 6 billion total bases of human genome), and that sequencer processes thousands of samples in parallel, so there is some economy of scale factors. It is sometimes very confusing how all these tech/techish companies have such bad financials.
Their financial reports consistently paint approximately 50% gross profit (revenue - cost of revenue). However, their R&D burn is enormous. Sales, marketing, general and admin costs are approximately equal to their gross profits.
In an alternate reality, you probably could have structured 23andme into a company making modest net profit. But that would not have matched the 6 billion price valuation either. And it's quite likely that a substantial part of the R&D burn was because they recognized the limitations of their core product.
Silicon valley's tokenomics are going bust because of the Fed's tightening. Loosing the reserve currency status is another matter altogether which should scare the future of the American economy.
That wouldn't matter. "Exorbitant privilege" is a thing but it's not that much of a thing. Notice Argentina, which has the opposite of our privilege, still gets loans - there's no other set of countries for lenders to go to.
The reason it looks like we have magic fiscal policy and Europe doesn't is that Europe just loves austerity and refuses to print money basically because they're stubborn. They could and should do it too.
It is interesting to me how their stated goals and planned future trajectory sounds like what happend to 'deCODE genetics' an Icelandic genetics company founded back in 1996 with a focus of "developing novel methods to identity, treat and prevent diseases"[0].
They really were on the forefront at that time, and have been an research powerhouse ever since[1] , but struggled for some time to find a sustainable revenue stream for their discoveries. Shortly after the turn of the millenium they came out with DNA diagnostic kits, that were able to diagnose various things known or thought to increase likelihood of various illnesses and conditions. And this product even came onto the market before 23andme.
But since 2012 (I think) they have been an independent research subsidiary of the US pharmaceutical company Amgen, where Amgen then can capitalize on the research output to make the drugs in question (and probably much more that we are not privy to).
I was an early adopter of 23andMe way back when they started. Before the privacy concerns. Before we had to worry about password uniqueness.
I didn’t renew my services with them. 1) I didn’t gain any value out of it, 2) their consistent blame of users that used recycled password for their breach.
I use ancestry.com as I’m working on dual citizenship with my family’s home country and find it worth the value. I was able to trace my family back through the Colonial Revolution all the way to the 1500s in Germany. That’s worth my money. Not 23andMe.
This is a damning indictment of Wojcicki's management of the company. Failed execution on the drug development strategy over 10+ years, lying about growth, pushing out the cofounder, never making a profit, hack that went undiscovered for months ... about the only area she's had success is raising money and that's in large part thanks to being a member of the Silicon Valley elite.
If she can't convince her family, friends, and connections to pony up more cash, the company will likely be sold as the DNA and profile information has real value. I worry that it will go to PE (like Ancestry.com) or a global data broker that will strip-mine the assets in every evil way you can imagine (and maybe even some we can't).
If they file for bankruptcy, they cannot legally destroy the data. The data is a salable asset, the law requires them to preserve it so that it can be sold and the money returned to creditors. And many of the promises they have given about it's use can and will be broken by the trustees dissolving the company.
The thing you need to consider when you give data to a company is not "will these people misuse this", but "once these guys go bankrupt and sell the data to the highest bidder, can it be misused?". See, for example, how if your company holds a large mailing list, it will typically be sold to a spam operation for a few bucks.
The minimum standard for giving serious personal information, like your DNA, to a company needs to be that the company receives a limited-time license to handle the data for the purposes they need to provide the service, and absolutely nothing more.
This makes me wonder if there could be strong enough laws around protecting genetic info (for example penalties that pierce the corporate veil) that those with fiduciary responsibility to the company could argue that destroying the data is really in the shareholders' best interest.
I'm completely uninformed in this area. Is there precedent for anything like this in the modern day?
Genetic information is PHI and is already covered under HIPAA.
The problem is that, like most companies, 23andMe isn't a HIPAA covered entity. Because they're not a healthcare provider or health insurer, it doesn't apply to them.
Genetic information is also controlled by other laws such as GINA. So it is already some of the most controlled health data there is. But that's not a very high bar in the US.
My mom, dad and sister have all done 23andMe, so it doesn't matter if I have or not. They have an entirely complete genome for me.
I agree with your limited license idea. It's just not ok that something like that can be dischargeable in bankruptcy. We don't have the ability to refuse consent in the first place, if our family provide it.
How can they link your family's DNA to you if they don't have a profile or account for you? How do they know you exist? And if they know you exist, how do they know you are related to that particular family without some DNA information about you to link to them?
OSInt will happily fill in the rest. Even if you don’t have a social media footprint, your family does and they shoved their contact list, with you in it, to Facebook 10 years ago.
That's a lot of information, but it is in no sense a complete genome. It does mean that someone who had that information could prove that your DNA belonged to a child of your mom and dad with high accuracy, or that you were a sibling of your sister. It also reveals that you don't have certain mutations, or that you do have a few.
But where your parents have different SNPs, there's no way to derive which of them you inherited. What you said is a bit like saying that, because you know all of the cards in a deck of playing cards, you know what hand someone is holding, except in a counterfactual world where there are 10,000 possible cards and you know that a deck only has 52 of them.
When people describe DNA as PII, I don't think they imagine being involuntarily cloned (yet..), I think they're imagining heritable conditions becoming public information through data leaks.
People have a right to privacy from imprecise yet correct information about themselves. Someone wouldn't want to explain an abusive parent to a prospective employer, but they could see a strong tendency to, schizophrenia, with DNA data leaks.
> The data is a salable asset, the law requires them to preserve it so that it can be sold and the money returned to creditors.
This would depend on the terms you agreed to at the time you signed the account and submitted your data. I'm aware a lot of companies like to bury marketing disclosures under a busload of T&Cs. But if you didn't agree having your marketable data allowed to be sold the sale the to creditors can't backdoor it in. They can only reach out to you to opt into new terms, which you should refuse.
My suspicion has always been that this sort of data–common SNPs from all comers and no specific phenotype collection–is not very useful for any sort of drug design. And if it is, publicly available data will be just as good as 23andMe's larger amount.
It was a bet worth trying, IMHO, but always a risky bet. And with bio, you don't find out if your bets work for a decade or more, because the iteration cycle is 10x slower than software for businesses/consumers.
My suspicion has been that SNP array data is not that useful for drug discovery. They’re targeting the most common/variable SNPs, which I suspect don’t have a large health effect (except for maybe late in life, otherwise how would they get passed down). I would suspect the more valuable targets would be rarer, or arise de novo (as is common in cancer, eg driver mutations).
The effect size of common SNPs is not informative about the effect of drugging their related genes. For example, the common variants near HMG-CoA reductase have very small but significant (confidently nonzero) effects. Yet drugging HMG-CoA reductase can reduce LDL cholesterol by ~40-50% (statins).
deCODE Genetics, whose history is very interesting and worth reading [1], was bought by Amgen based on this premise.
Note, however, that SNPs like the one you pointed out are relatively infrequent. Amgen was expecting a two digit % improvement in their pharma pipeline by using GWAS insights.
> Note, however, that SNPs like the one you pointed out are relatively infrequent.
If you mean that SNPs with small effect sizes don't always point to useful drug targets with big druggable effects, that is possible, but this remains an open question and is the subject of intensive research right now.
Yes, I agree. The trick is probably to find cell-specific SNPs located in regulatory regions so that there are no off-target effects. Massive screens using single-cell perturbations will help to gain some insights.
You’re right that rarer SNPs have big effect sizes. However the problem is that they’re rare so you don’t make much money treating ppl with rare diseases.
They could help in stratifying patients and strategically targeted therapy. They can't be easily targeted (drugged) drirectly but they could lead to therapeutics if I understand correctly.
Well, nobody is really interested in the test to find out they have G at SNP 12213244243 either. They do it specifically to get the ancestry, health, and potential physical bias type information you mention - that's the only real data anyone is talking about when referencing the test. Plenty of companies would love to have that kind of data to target users with.
The real killer would probably be around how useless healthcare regulation could make having most of that data. I'm not sure how far that really extends or what you manage to sign away by sending in your sample though.
That's kind of what I mean by it not being worth much though. Sure, leaks are embarrassing - the most common issue here is finding out that your dad isn't your real dad.
But it's not useful for evil purposes because it's not /accurate/. It doesn't reveal all that much real additional information about your ancestry and health!
Since it's decided before you're born:
1. its effects are mostly already expressed and people can know them either by looking at you or (like life insurance) by making you get a medical exam before buying their services.
2. it can get out of date for cultural reasons. Your biological ancestry isn't always how you behave, like if you're adopted. Whereas last name + zip code is a pretty good predictor already.
But 23AndMe does have a lot of that information about their customers. The reason they have it is because they asked for it in surveys.
> But it's not useful for evil purposes because it's not /accurate/. It doesn't reveal all that much real additional information about your ancestry and health!
I'm not actually convinced that accuracy is all that important to the value of personal data. Or perhaps I should say I don't think think there's an accuracy floor at which the value drops to $0.
Not in the field, but i would assume the data is combined to create profiles to be used to infer data about other people with similar data, so now they have dna into the profiling mix. And in combined I mean the rest of your digital footprint from other sources, not a survey on 23
Strongly disagree. Through 23andme, I was able to confirm that I have the mutation for the MTHFR gene, which explained why my vitamin B12 always ran low. Once I started supplementing with the methylated form of B12 and folic acid, I noticed a huge improvement in how I felt vs when I used cyanocobalamin for B12. Not to mention some very interesting side effects, like improved tolerance for caffeine.
I had this same issue. I bought Pure Encapsulations Adenosyl/Hydroxy B12 and Thorne Research - 5-MTHF 1 mg Folate from amazon but never felt any different.
Was my B12 a dud, does everyone with this marker see improved results after supplements?
The health data is pretty underwhelming for most people, outside of a few highly studied genes like BRCA variants. I did a genetic test and it reported that according to one study I had something like a 3% risk of a shoulder injury which is higher than the population average of 2%. So what am I supposed to do with that information, wear shoulder pads or something?
Due to GINA, medical insurance companies aren't allowed to use that data for setting rates or denying coverage so it's useless to them. In theory life insurance companies could use it, but at the scale they operate it's not worth the cost.
Isn't there significant value in the medical information, deciding who to offer insurance to and who to refuse to cover in the event that Republicans gain control of the government and repeal the Affordable Care Act?
But is it bad for there to be a cache of data for research? I don't even care if they make a great profit because if there's a great profit to be had it means that some drug came out of the research. While I'm generally cynical about many of the drugs, I do like the fact that the industry tries hard. Eventually they'll make something I want to take. And if this data helps them, all the better.
There are 3-4 companies like Invitae (and not 23andMe) that genetic counselors send their tests to for analysis and treatment coordination with the MD.
Like another comment said, I think the DNA data 23andMe gets from their swabs and analysis is likely inferior from a drug-dev standpoint vs. what the pro firms do linked with the genetic counseling teams.
In that light, 23andMe's options don't look good. However, probably good news for the patient. As I was told by a genetic counselor at one of the main hospital systems that do this, a bulk of their work is rolling back expectations from patients who did 23AM first, and then when to a GC.
Unending confidence and carefully curated reality distortion fields. Having watched the CEO at a previous job do it, it's infuriating and a gigantic pain in the ass for engineers to have a leader lie to investors and then have to deal with the fallout as deadlines approach.
It's all going to come out in discovery. She used to work for a hedge fund. Data collection and disregard for other peoples' privacy runs in the family.
Not encrypting user data and then being hacked resulting in compromised user DNA details certainly isn't a positive for PR. Can someone tell me why more companies don't encrypt data at rest and if that would have prevented the leak?
Because actually encrypting at rest usually offers a ton of hassle for little actual additional security.
For example, if done right, you need to decrypt with some form of passcode/system on boot, etc, etc. But once "unlocked" the system has access to the data, and usually the hackers come in via the "live" system.
Things like backups are usually encrypted but not always, but the number of incidents of people stealing physical media is pretty low compared to ransomware/remote hacks.
Encrypting data at rest in a data center doesn't provide as much protection as you might think. The level of system access in order to access the raw data files is usually about the same as accessing the higher level database system itself. So at best you're adding a couple of extra steps to the attack.
Since there are usually significant costs to encrypting production databases, such as CPU utilization, many organizations choose to focus on other security controls. Data encryption is just one of many security controls at play in a scenario such as this and the security team has to carefully pick their battles to avoid just being completely ignored as the "say no to everything" guys.
Now we get to see the compromises they choose when desperation sinks in.
I think getting the results of one of these tests would be interested, but I haven’t been able to bring myself to do it, due to trust issues related to handing over something like that to a company with an unproven track record when it comes to privacy and integrity. I figured it was only a matter of time before they were hacked or they decided to sell out their users for profit. They’ve already been hacked, so now it’s time to see what they willingly do with that data when their back is against the wall.
I wish I could get a test done by a lab, be given the results, and have that be the end of it. No accounts, no forever records kept by the lab/company. Just a simple testing service that isn’t trying to be a tech company.
> I wish I could get a test done by a lab, be given the results, and have that be the end of it. No accounts, no forever records kept by the lab/company. Just a simple testing service that isn’t trying to be a tech company.
EMR has put paid to this pretty handily. Every positive STD test anyone ever gets from now on is yours for the taking if you can crack LabCorp's database.
“Entrepreneurs are not at all like ordinary businessmen. An entrepreneur who is not in trouble closes no avenues, keeps a lot of balls in the air, and will never tell you the whole truth when a half-truth will do. An entrepreneur who is in trouble will lie, cheat, and steal. He will smuggle cocaine or ship bricks. We should never measure an entrepreneur by the standards of a rock-solid businessman.” — attributed to Kenneth Rind[1]
>I guess it depends on what he meant by “rock-solid.”
Exactly.
There's more than one kind of rock-solid.
Someone who is a rock-solid entrepreneur will not lie, cheat, or steal whether they are in trouble or not.
OTOH an "ordinary" non-entrepreneurial businessman may just be one who has not faced trouble (may also have a lot less risk of that) and not started to lie, cheat, or steal. Yet. But may do so any minute at the drop of a hat and you would never know.
No need for a difference in codes of ethics, rather different kinds of rock-solid for thee than me.
It's easy to remember when Enron was a rock-solid client, packed with ordinary businessmen from ordinary business schools and typical backgrounds, until one day it wasn't. Wasn't rock-solid that is, and able to pay. All the businessmen there remained ordinary after that regardless, except for the few whose ethical transgressions were among the "unavoidable handful" for which there were serious legal implications.
The sentence immediately preceding your quote are things that a "rock-solid businessman" would never do according to Rind. So it sounds like he is saying that theoretical person has a higher code of ethics for whatever reason.
Because they're intending to be around for the next decade or more, they need to preserve their appearance of probity, and so they'll do fewer things which would make them look bad.
Rock-solid businessmen understand business and how it works. Nothing personal, only business.
Entrepreneurs are strongly driven by ego, their identity is personally tied to the thing they’re doing. There is immense pressure to succeed, not just for the money, but for their personal validation. It’s not enough to succeed, people must see you succeed.
Otherwise, they’d just open some boring small business no one cares about instead of a hot new startup.
> I haven’t been able to bring myself to do it, due to trust issues related to handing over something like that
Same here. Imagine the value of this data in the hands of insurance companies that are willing to misuse it. Will we, as a society, be able to prevent that "value" from being realized?
Federally, health insurers cannot use this data under GINA[0] and SB41[1] in California requires explicit permission be granted by the customer in order for the data to be shared with 3rd parties. In general, however, insurance companies are going to be reluctant to trust this information as it’s not established that it has real medical value outside the handful of markers we’ve clearly identified.
Oh, it's against the law! Well that's ok then, companies never break laws.
In reality, there will be (and probably already is) an impenetrable web of "data aggregators" buying genome data from other aggregators who are buying it from still other aggregators, so that the info will be thoroughly laundered by the time it actually gets to the insurance company to hike your premiums. This both obscures its true origin and lets there be a little contractor to throw under the bus when discovered and deflect blame. Don't fucking kid yourself: if personal genome data is useful, it will be used, no matter what the law says.
I'm more worried about a scenario like Elon Musk buying the company for the data and making the data available to extremists who would love to target people of certain ethnic backgrounds who happen to be in positions of power.
281 comments
[ 1.9 ms ] story [ 221 ms ] thread> Sequoia, which invested $145 million in 23andMe, still holds all its shares, he said. Today they are worth $18 million.
Interesting that they weren't forced to liquidate/distribute the shares to their LPs
I smell a conspiracy theory.
1. your DNA results are not interesting to anyone even if they're evil.
2. if someone wanted your DNA in particular, they just need a bit of your hair, so there is already absolutely nothing you could do to stop them.
I would imagine that any fund is allowed to make losses on specific investments, least of all a hyper-aggressive VC fund which will create losses on many (most) positions, by design. And then once losses happen they should be treated for what they are - just another position where decisions should be made based on the outlook, and not based on the history.
"Sequoia is abandoning the 10-year venture fund, in which limited partners, the outside investors that contribute to the fund, expect to get paid back over a decade. The firm said it’s establishing a single fund, the Sequoia Fund, that will raise money from LPs and then funnel that capital down to a series of smaller funds that invest by stage.
Proceeds from those funds will feed back into the Sequoia Fund. With no time horizon, Sequoia can hold public stock for longer stretches, rather than distributing those shares to LPs. Investors who want liquidity can pull money out instead of waiting for distributions."
https://www.cnbc.com/2021/10/26/sequoia-changes-fund-structu...
The other had an idea.
Make sure you are the co-founder with the sustainable advantage
but more of an object lesson that as the needs of the company change, the board/your co-founders will constantly re-evaluate whether you are still adding value.
Looking at those financials... yikes. I am kind of shocked how poorly they are doing.
Their huge problem is hackers. All the latest news is how hackers have grabbed up those databases.
What's likely to happen, they'll self restructure, get to profitability, start a stock buyback program. But IF and WHEN?
The consumer side is clearly struggling because of the problems mentioned in the article (they only need one test in their life, public perception is bad because their security has had breaches). So this needs a pivot where you can change the public's perception from a one-time test to continuous health monitoring through blood markers or something similar, expand to tests other than genetic and make it a repeatable, accurate test that gives you more information (and obviously stop leaking people's data).
But why not focus on the B2B side? Sell access to their databases. I'm sure computational biology and/or pharma companies need this information. It makes sense to do vertical integration by manufacturing your own drugs, but not for a cash-strapped business that has little incoming revenue to sustain further development. Presumably they are selling their genetic data, but I don't get why it's not giving them a revenue stream. Let GSK manufacture the drugs using your genetic info, with a profit share for any drugs made this way. They mention a collaboration with GSK in the article, but why was this stopped?
"Stop leaking data"? Sounds like a step forward.
> But why not focus on the B2B side? Sell access to their databases.
So, it's actually: try to sell your customers' data instead of leaking it? :O
Sorry if that sounds snarky, but are we sure there are enough customers who want to pay to give their data to a company so that company can immediately sell their data on to other companies?
Are you sure?
I mean, presumably there are different types of DNA testing. Doesn't 23andme run basically the cheapest test they can get away with? A user can't tell if the test measured 16 bytes or 1.6 gigabytes of genetic information, and if I was trying to launch a consumer DNA geneology service, I'd want to get network effects, so I'd want a test that was very easily affordable.
Who says their records are thorough enough to be valuable to drug companies?
No. That's part of why they're test is so expensive and their financials are so lackluster. They are apparently using a customized version of the lllumina's Global Screening Array according to their website and several other sources that show up in search results. That's a legit research quality genotyping platform from a world leading laboratory in the genetics space. This post from 2020 has a decent high level overview about it in the context of 23&me [0] though it might be slightly outdated by now and I've never heard of the company (xcode?) that wrote it (nor did I bother to look at what their product is.)
> If I was trying to launch a consumer DNA geneology service, I'd want to get network effects, so I'd want a test that was very easily affordable.
That's a pretty neive perspective. It ignores the value propositions of 23&me's product, economies of scale in the direct to consumer genetic testing space which were in large part enabled by 23&me's success, and all of the thorny bits related to questions about accuracy when presenting results. Not to mention that 23&me certainly capitalized on the network effect (which is being criticized a fair amount in this thread.)
> Who says their records are thorough enough to be valuable to drug companies?
GSK. See my other comment in this thread [1]
[0] https://www.xcode.life/23andme/23andme-v5-chip-dna-raw-data-...
[1] https://news.ycombinator.com/item?id=39202583#39204621
In 2018 GSK made a $300M equity investment in 23andMe as part of a 4 year collaboration (with the option to extend for a fifth year) under which GSK had exclusive access to their data for use in drug target discovery programs, but [0]:
> All activities within the collaboration will initially be co-funded (50%/50%), with either company having certain rights to reduce its funding share for any collaboration programme.
So it seems they not only lost out on 5 years of developing their B2B business, but committed to covering a portion of the R&D costs over that period as well. There were terms about profit sharing on new developments, so it was a bet.
It doesn't sound like it worked out quite as well as either sided hoped though because in October 2023 (after the 5 year agreement) they entered into another agreement but this time [1]:
> Under an amendment to their Collaboration Agreement, 23andMe will receive a $20 million upfront payment for a one year, non-exclusive data license. > [...] > for a 12-month period, and [23andMe will] offer its research services for analyses of the data over that same period. Any new drug discovery programs that GSK chooses to initiate during the agreement will be owned and advanced solely by GSK.
[0]: https://www.gsk.com/en-gb/media/press-releases/gsk-and-23and...
[1]: https://investors.23andme.com/news-releases/news-release-det...
The company hasn't the funds (or time) to develop drugs and I'm not sure why any investors would invest in them rather than a company with a track record of doing so or a potential drug.
Selling their customers data to other companies (or partnering with) that can create drugs from it would seem the logical choice, along with "subscription health reports" and anything tacked onto the DNA/blood profiling/testing product to keep the data coming in.
Attempting to discover drugs is a proven way to make a small fortune by starting with a large fortune. Everything north of the Charles River in Boston would vanish if people stopped believing though.
Was there an issue with targeting sub-populations for clinical trials beforehand?
At an ELI5 level, if you're hoping your drug candidate will help cure disease X, you sign up patients with disease X to join your clinical trial. That's not the hard part!
(source: family member works at $bigPharma)
a) If you suspect there is a significant pharmacogenetics component to what you are studying (or related to disease progression).
b) You're working on something preventative.
From a previous job I worked at (we made PCR tests), there was interest on screening for APOE genotypes to enrich an Alzheimer's drug trial - drug maker believed that APOE genotype would have a significant impact on drug performance.
But you're absolutely right that you can often (usually?) do 'enough' enriching without genetic information.
An example of this is when they tested the covid vaccine, they wanted to make sure they had enough participation from African American and other ethnicities since these were usually under represented in clinical trials relative to the population, and there are sometimes subtle variations in the way peoples body respond to drugs depending on their race.
To find new targets yes but to develop a drug to market (and revenue/profit) takes 10-15 years. Even a drug to a state that gets approved for human testing seems beyond what they could do themselves without so much funding it would wipe out the existing shareholders. Selling to/Partnering with other drug companies with targets they've discovered/IP from DNA etc would seem more realistic.
Perhaps ok for a startup, not ok for a many years old established company.
In the words of Boromir: "One does not just become a drug company."
Haven't they already been trying this ... for years?
(Way back in 2018): "GSK and 23andMe sign agreement to leverage genetic insights for the development of novel medicines"
https://www.gsk.com/en-gb/media/press-releases/gsk-and-23and...
Why would it be surprising to value the latter kind of business higher than the former?
If you either create a brand new market, or are taking on a drastically undeserved market... you absolutely need to be growing faster than inflation. Or your brand new company would be worth $1 the first year, $1.03 the second year..
Even outpacing inflation by 1.2 - 2 times. It's not insane growth, but it is consistent. Better yet is that they tend to have a decent number of assets too.
> It’s really hard to grow a business at 10 or 500 times that rate.
How often is 10 to 500 times growth seen in reality? I think most start-ups fail before the second year.
> Why would it be surprising to value the latter kind of business higher than the former?
Highly stable and proven growth as opposed to highly unstable and unreliable growth.
The small business is successful and so the calculus as to what it is worth is much more explicit.
The extreme example of this is commercial real estate with a long-term tenant (think: the person who owns the land a post office is sitting on). Everything is known (lease length, amounts, current interest rates, etc, etc) that the value/sale price is a simple formula.
I get new updated health information from 23andme perhaps 2 times per year. It's mind boggling to imagine that with all the research being done that is all that makes the cut as interesting and trsutworthy enough for 23andme to tell me about it.
In other words, a majorly failed content strategy based on fear of being shut down for providing insufficiently vetted health info.
Meanwhile, someone created multiple accounts and the ancestry results didn't match even when collected from the same person.
I have some pills for you. Just for you, mind you. Special made. Sign up for this subscription and we’ll send you a new supply each day…
I'm not looking for a single source of authority and I will not be disappointed if an updated "chip" or study reverses a previous recommendation/insight.
23andme had a blog post about the MTHFR gene and why they don't include it. I'm not there to read the blog I want to know interesting things about my own genetics, even if what is interesting is that they don't think it matters! This is just bad content strategy even if it's justifiable from a "healthcare" perspective.
Why so much worry about a customer getting the wrong impression from one of the health reports? Seems a bit absurd. How much money did it cost to put in place all the warnings about the potentially upsetting impact of various mutations.
I wish that 23andme had focused on being an entertaining way for people to discover more interesting stuff about themselves, rather than a rarefied place where only the most well-vetted research is acknowledged to exist.
If a healthcare startup get shout-outs from Oprah, SNL and branded Barbie dolls, is that a bad sign?
I don’t understand how anyone ever thought 23andMe had a viable business model selling a test that only needed to be done once(and apparently selling it at a loss to boot)
There's a reason that these complex undertakings are made by boring conglomerate companies that have decades of experience in the industry, produce many different products in that sector, and whose CEOS don't try to get on the covers of magazines.
That is a pretty awful click-bait title. Presumably the "$0" is in reference to their sub $1 share price?
The fact that they put everything behind a paywall _and_ slap on click-bait titles is confusing to me. I can't imagine paying for access to that type of writing.
23andMe never had a business plan, but it didn't matter because their investors figured they had the connections to make it work (look at the explanation of how they booted out the other founder). Their Libertarian/lazy attitude towards regulation is just an example of how couldn't be bothered to even do the necessary homework. Their attitude towards their data breach is another.
Makes me wonder how good their science is. I don't know anyone on that side of their business, but comments from friends who used to work their about how their parts worked are not encouraging.
I was a heavy believer of 23andMe until this point. I answered all of the available research questions, which was a thing that took absolutely hours and was filled with semi-invasive medical questions. I did this under the premise that I would hopefully be helping research and I felt really rewarded having completed all of them. Then, they dropped the + bombshell and I felt really rugpulled. I paid them for genotyping on their v4 and v5 platforms -- so I paid twice, I referred friends, I bought people kits, I helped research...and now I was being asked to pay a subscription for what I was promised to begin with? Eesh.
I think the fundamental issue is that when the FDA stepped in and told 23andMe to cut back on their reporting that they really hit a roadblock. Promethease always gave better results than 23andMe, but when that happened it commoditized genotyping fully. Just download your data from any provider and use a third-party tool and you're set.
Fast forward to 2024 and they're stuck with a platform that's semi-limited and they haven't delivered on any of the research deliverables that many people wanted in the first place. The idea of getting new genetic reports monthly was appealing and simply never materialized. It's no surprise people aren't hot on this as a business -- but what is surprising is that they were completely passed by startups like Nebula Genomics in offering whole-genome sequencing and competitive data access. I think the stored data they have is their only advantage, but they don't seem to know how to leverage it.
I really wonder what happens to that data if they get acquired or shut down. The possible new owner of the data might have a completely different business case in mind. This scenario is something you usually don't consider when you give your data to a company that you trust at the moment.
Which is to say: No they wouldn't just do that, it would cut their business too much despite being a way to make more money per client.
But getting it from a company that already has it avoids that downside.
This scenario is something you usually don't consider when you give your data to a company that you trust at the moment.
Why would you not consider it?
But for the same reason I'm just a one time customer, all I needed was for 23andme to give me a very expensive hospital test at a 95% discount, and I didn't need them anumore
https://yourdnaportal.com/advanced_ancestry_analysis
This is inescapable; genetics usually can't show causation, for instance because you can't do an experiment where you change someone's genes.
Geneticists seem to deal with this by using statistics like GWAS that are obviously just correlation, adding a sentence that correlation doesn't show causation, and then just proceeding and pretending like it does.
23andme went with the "only give very limited information" which I respect but is ultimately pretty much useless compared to just running your data against dbSNP and telling me that I have this or that marker for this or that disease backed by this or that paper.
Though if you find whatever gene is most correlated with something, what are the options for it not to be causation? If the chance of causation is high enough, it makes sense to proceed as if the risk is real.
There's a more advanced form of being bad at this where you think you can show causation by controlling for everything in the environment. This is also wrong; it produces something called collider bias.
> Though if you find whatever gene is most correlated with something, what are the options for it not to be causation?
1. It's a coincidence and it's never causal.
Imagine an OSS project releases a bugfix and you diff the old and new versions. (This is basically GWAS.)
The bugfix part of the diff caused it to be fixed. The updates to the copyright dates or changelogs didn't.
2. It's causal, but the causal chain involves a specific environmental factor, and we should change that instead.
For instance, you can say every human has a genetic disease that prevents them from producing their own vitamin C, which most other mammals can do. But instead of calling scurvy a genetic disease we just eat fruits and vegetables.
Half the point of analyzing statistics is to filter out coincidences, and that applies to correlations too. If something is a coincidence, it won't hold up as a proper correlation under reasonable amounts of analysis. So when the premise is we're starting with correlations, I think it's alright to assume they're mostly not coincidences.
> The bugfix part of the diff caused it to be fixed. The updates to the copyright dates or changelogs didn't.
In that case the bugfix is the "gene most correlated", isn't it? Give it a few generations to randomly spread, and the signal will be far stronger on the bugfix gene than on the copyright gene. (And if it hasn't been spreading for generations then you won't have enough samples to find either gene.)
> 2. It's causal, but the causal chain involves a specific environmental factor, and we should change that instead.
> For instance, you can say every human has a genetic disease that prevents them from producing their own vitamin C, which most other mammals can do. But instead of calling scurvy a genetic disease we just eat fruits and vegetables.
If you're testing just humans, you'll get a 0% correlation because everyone has that gene.
If you're testing across mammals, then "WARNING: Prone to scurvy". Which is a completely correct and causal result about a genetic problem, with easily accessible treatments.
So, I don't understand your example at all.
Have maybe a small ounce of humility in this respect.
Thousands of very smart people do a lot of dumb things. People still work on string theory. The people working on proper causal inference are also smart, work hard, and have Nobel Prizes.
https://www.nobelprize.org/prizes/economic-sciences/2021/pop...
On the other hand, the people doing genetics who aren't careful about it produced the 23AndMe report which says I have a "17% chance of having green eyes". Against what counterfactual?
> In fact, you can correct for the exact problem you describe with sibling studies where people have the same environment but different genes.
"Correcting" is a wrong way to think about it. Generally speaking, overcorrection is worse than undercorrection because of collider bias. You need to choose a study design that's correct in the first place.
You're describing a natural experiment, which is better than a GWAS of self-selected 23AndMe customers, but does have problems (silly one: selection bias because the sample only includes people with siblings) and more importantly is only guaranteed in the study environment (eg people who live in the UK in 2008) but gets reported without identifying what that environment is.
> You need to choose a study design that's correct in the first place.
That's precisely what a sibling study is: a study design that's correct in the first place, as far as the problem we're discussing goes.
> You're describing a natural experiment
That is correct: a sibling study, which corrects for the problem we're discussing, often contains aspects of a natural experiment, namely that the siblings tend to naturally be siblings, rather than raised together purely for the purpose of a study :)
Apparently 23andMe doesn't let you download the data anymore, I just tried. I wanted to give Promethease a go, seems interesting.
(1) Is that being associated with a penis pill pusher is brand destruction. Any normal person would get this right away without trying but executives just can’t see what their actions look like to their customers.
(2) It’s a form of achievement laundering. Lemonaid health failed, pure and simple. A circle jerk acquisition is alchemy that turns a failed business into an apparent success, now the founders, management and people who invested in it will get accolades for being successful when really the Silicon Valley Mafia stole from public investors to maintain its image.
It ought to be a lot harder for businesses to get acquired. Maybe we need a vote by the customers, maybe when Google buys a company they should be required to spend a 1/4 of the money on ads warning people that the business could get shut down, etc.
[0] https://youtu.be/C3T_IaIGQxo?t=141
In an alternate reality, you probably could have structured 23andme into a company making modest net profit. But that would not have matched the 6 billion price valuation either. And it's quite likely that a substantial part of the R&D burn was because they recognized the limitations of their core product.
https://investors.23andme.com/news-releases/news-release-det...
I give it 10 years time before it turns into a town just living on its old glory, with tourist attractions like:
- "This used to be a bar where Google's exec team drank"
or
- "This was the hotel Mark Zuckerberg threw an Oculus Rift into the pool"
or
- "Oracle datacentre re-enactment theme park, with real old beard Oracle DBA you can interact with"
Reference: https://youtu.be/AUrHVWa3s2Y
The reason it looks like we have magic fiscal policy and Europe doesn't is that Europe just loves austerity and refuses to print money basically because they're stubborn. They could and should do it too.
They really were on the forefront at that time, and have been an research powerhouse ever since[1] , but struggled for some time to find a sustainable revenue stream for their discoveries. Shortly after the turn of the millenium they came out with DNA diagnostic kits, that were able to diagnose various things known or thought to increase likelihood of various illnesses and conditions. And this product even came onto the market before 23andme.
But since 2012 (I think) they have been an independent research subsidiary of the US pharmaceutical company Amgen, where Amgen then can capitalize on the research output to make the drugs in question (and probably much more that we are not privy to).
[0]: https://en.wikipedia.org/wiki/DeCODE_genetics
[1]: https://www.decode.com/publications/
edit: wording and spelling
I didn’t renew my services with them. 1) I didn’t gain any value out of it, 2) their consistent blame of users that used recycled password for their breach.
I use ancestry.com as I’m working on dual citizenship with my family’s home country and find it worth the value. I was able to trace my family back through the Colonial Revolution all the way to the 1500s in Germany. That’s worth my money. Not 23andMe.
If she can't convince her family, friends, and connections to pony up more cash, the company will likely be sold as the DNA and profile information has real value. I worry that it will go to PE (like Ancestry.com) or a global data broker that will strip-mine the assets in every evil way you can imagine (and maybe even some we can't).
The thing you need to consider when you give data to a company is not "will these people misuse this", but "once these guys go bankrupt and sell the data to the highest bidder, can it be misused?". See, for example, how if your company holds a large mailing list, it will typically be sold to a spam operation for a few bucks.
The minimum standard for giving serious personal information, like your DNA, to a company needs to be that the company receives a limited-time license to handle the data for the purposes they need to provide the service, and absolutely nothing more.
I'm completely uninformed in this area. Is there precedent for anything like this in the modern day?
Genetic information is also controlled by other laws such as GINA. So it is already some of the most controlled health data there is. But that's not a very high bar in the US.
It's not even really that unreasonable, if they want to advertise that they can provide health information, they don't get to pretend that they don't.
I agree with your limited license idea. It's just not ok that something like that can be dischargeable in bankruptcy. We don't have the ability to refuse consent in the first place, if our family provide it.
But where your parents have different SNPs, there's no way to derive which of them you inherited. What you said is a bit like saying that, because you know all of the cards in a deck of playing cards, you know what hand someone is holding, except in a counterfactual world where there are 10,000 possible cards and you know that a deck only has 52 of them.
People have a right to privacy from imprecise yet correct information about themselves. Someone wouldn't want to explain an abusive parent to a prospective employer, but they could see a strong tendency to, schizophrenia, with DNA data leaks.
This would depend on the terms you agreed to at the time you signed the account and submitted your data. I'm aware a lot of companies like to bury marketing disclosures under a busload of T&Cs. But if you didn't agree having your marketable data allowed to be sold the sale the to creditors can't backdoor it in. They can only reach out to you to opt into new terms, which you should refuse.
It was a bet worth trying, IMHO, but always a risky bet. And with bio, you don't find out if your bets work for a decade or more, because the iteration cycle is 10x slower than software for businesses/consumers.
Note, however, that SNPs like the one you pointed out are relatively infrequent. Amgen was expecting a two digit % improvement in their pharma pipeline by using GWAS insights.
[1] https://en.wikipedia.org/wiki/DeCODE_genetics
If you mean that SNPs with small effect sizes don't always point to useful drug targets with big druggable effects, that is possible, but this remains an open question and is the subject of intensive research right now.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8129253/
They could help in stratifying patients and strategically targeted therapy. They can't be easily targeted (drugged) drirectly but they could lead to therapeutics if I understand correctly.
I'd be nice if they could be prevented from doing that, possibly by the court system.
(They do have valuable data, but it's the survey answers and friend and family connections. Nobody really cares you have a G at SNP 12213244243.)
The real killer would probably be around how useless healthcare regulation could make having most of that data. I'm not sure how far that really extends or what you manage to sign away by sending in your sample though.
But it's not useful for evil purposes because it's not /accurate/. It doesn't reveal all that much real additional information about your ancestry and health!
Since it's decided before you're born:
1. its effects are mostly already expressed and people can know them either by looking at you or (like life insurance) by making you get a medical exam before buying their services.
2. it can get out of date for cultural reasons. Your biological ancestry isn't always how you behave, like if you're adopted. Whereas last name + zip code is a pretty good predictor already.
But 23AndMe does have a lot of that information about their customers. The reason they have it is because they asked for it in surveys.
I'm not actually convinced that accuracy is all that important to the value of personal data. Or perhaps I should say I don't think think there's an accuracy floor at which the value drops to $0.
Was my B12 a dud, does everyone with this marker see improved results after supplements?
After all, they didn't need your genes when they were denying care in the 90s.
Like another comment said, I think the DNA data 23andMe gets from their swabs and analysis is likely inferior from a drug-dev standpoint vs. what the pro firms do linked with the genetic counseling teams.
In that light, 23andMe's options don't look good. However, probably good news for the patient. As I was told by a genetic counselor at one of the main hospital systems that do this, a bulk of their work is rolling back expectations from patients who did 23AM first, and then when to a GC.
I mean she was leader of YouTube. Why wouldn't she be able to develop a drug development strategy based on genomics?
But hey, rich people can do anything right?
Or maybe Palantir.
For example, if done right, you need to decrypt with some form of passcode/system on boot, etc, etc. But once "unlocked" the system has access to the data, and usually the hackers come in via the "live" system.
Things like backups are usually encrypted but not always, but the number of incidents of people stealing physical media is pretty low compared to ransomware/remote hacks.
Since there are usually significant costs to encrypting production databases, such as CPU utilization, many organizations choose to focus on other security controls. Data encryption is just one of many security controls at play in a scenario such as this and the security team has to carefully pick their battles to avoid just being completely ignored as the "say no to everything" guys.
I think getting the results of one of these tests would be interested, but I haven’t been able to bring myself to do it, due to trust issues related to handing over something like that to a company with an unproven track record when it comes to privacy and integrity. I figured it was only a matter of time before they were hacked or they decided to sell out their users for profit. They’ve already been hacked, so now it’s time to see what they willingly do with that data when their back is against the wall.
I wish I could get a test done by a lab, be given the results, and have that be the end of it. No accounts, no forever records kept by the lab/company. Just a simple testing service that isn’t trying to be a tech company.
EMR has put paid to this pretty handily. Every positive STD test anyone ever gets from now on is yours for the taking if you can crack LabCorp's database.
“Entrepreneurs are not at all like ordinary businessmen. An entrepreneur who is not in trouble closes no avenues, keeps a lot of balls in the air, and will never tell you the whole truth when a half-truth will do. An entrepreneur who is in trouble will lie, cheat, and steal. He will smuggle cocaine or ship bricks. We should never measure an entrepreneur by the standards of a rock-solid businessman.” — attributed to Kenneth Rind[1]
[0]: https://news.ycombinator.com/item?id=28060166
[1]: https://books.google.com/books?id=MQvGc8Ee1SsC&pg=PA137&lpg=...
> "We should never measure an entrepreneur by the standards of a rock-solid businessman".
Is the author saying that a rock-solid businessman follows a higher code of ethics?
Exactly.
There's more than one kind of rock-solid.
Someone who is a rock-solid entrepreneur will not lie, cheat, or steal whether they are in trouble or not.
OTOH an "ordinary" non-entrepreneurial businessman may just be one who has not faced trouble (may also have a lot less risk of that) and not started to lie, cheat, or steal. Yet. But may do so any minute at the drop of a hat and you would never know.
No need for a difference in codes of ethics, rather different kinds of rock-solid for thee than me.
It's easy to remember when Enron was a rock-solid client, packed with ordinary businessmen from ordinary business schools and typical backgrounds, until one day it wasn't. Wasn't rock-solid that is, and able to pay. All the businessmen there remained ordinary after that regardless, except for the few whose ethical transgressions were among the "unavoidable handful" for which there were serious legal implications.
Entrepreneurs are strongly driven by ego, their identity is personally tied to the thing they’re doing. There is immense pressure to succeed, not just for the money, but for their personal validation. It’s not enough to succeed, people must see you succeed.
Otherwise, they’d just open some boring small business no one cares about instead of a hot new startup.
Same here. Imagine the value of this data in the hands of insurance companies that are willing to misuse it. Will we, as a society, be able to prevent that "value" from being realized?
[0]: https://medlineplus.gov/genetics/understanding/dtcgenetictes...
[1]: https://leginfo.legislature.ca.gov/faces/billTextClient.xhtm...
In reality, there will be (and probably already is) an impenetrable web of "data aggregators" buying genome data from other aggregators who are buying it from still other aggregators, so that the info will be thoroughly laundered by the time it actually gets to the insurance company to hike your premiums. This both obscures its true origin and lets there be a little contractor to throw under the bus when discovered and deflect blame. Don't fucking kid yourself: if personal genome data is useful, it will be used, no matter what the law says.