It is definitely the case that the parents of babies and kids with life-threatening conditions are often one the most motivated people you see on Earth and what they accomplish sometimes is truly incredible. My heart goes out to them including the OP - I can only imagine how hard it must be....
While I am truly grateful for him and the team for their contributions to neonatal genetics (and hosting me in San Diego for a few days to show me how I could help), Rady was actually the unnamed lab that failed to diagnosis my son.
And this happens all the time. The WGS NICU diagnostic rate is only ~30%, depending on who you ask. Just because people have been working at this for a decade and products exists, doesn't mean it's a solved problem.
I don't know if you read until the end of my post, but I did run a small experiment in collaboration with an academic geneticist and outperformed the first-line clinical labs across the board. My approach, which is essentially Claude Code for genetics, is fundamentally different and novel than how this work is done today and seems to perform much better in early experiments. Time will tell is this generalizes to all clinical work.
I'm planning on publishing evals and benchmarks in the next few weeks, but out-of-the-box systems actually don't do very well for a variety of reasons.
> The WGS NICU diagnostic rate is only ~30%, depending on who you ask.
Agreed. It does not automatically mean, however, that it can be significantly improved with better variant interpetation or better analysis of the same wgs data in general sense.
> I'm planning on publishing evals and benchmarks in the next few weeks, but out-of-the-box systems actually don't do very well for a variety of reasons.
Happy to see it. I wish you all the luck and will be the first one praising your solution if I see convincing results.
I wouldn't say anything is automatic or taken for granted, but it is actually relatively common for more thorough reanalysis to uncover something that the first pass missed. I hinted at this in the post, but the reason that this doesn't happen today is human bandwidth.
A core part of my thesis is that that this highly specialized human bandwidth can be scaled with AI.
It may work. It may not work. But I would feel bad if I didn't give it a try.
> Happy to see it. I wish you all the luck and will be the first one praising your solution if I see convincing results.
Appreciate that! Hopefully, they will come.
this is a very common reaction to people doing things with llms and i think the effects of it can be somewhat insidious. you constantly see people out there vibing their way to something that has already been discovered somewhere else, but they didn't know that and in many cases wouldn't have known how to find that thing even if they did.
the framing of "the llm tricked you into thinking you discovered something" while technically true in many cases, very strangely casts the positive outcome of a person being linked in a very engaged manner to something they wouldn't otherwise know or found out into something to be looked down on, and sort of just discourages people from trying stuff themselves that wouldn't be possible for them without something like an agent. it's okay if someone else already found the thing. for areas like science and research, it's actually a good thing if something you did repeated the work of someone else. it validates the original piece of work, and it tells you the things you were trying were on the right track to begin with.
They have to be useful, otherwise nobody comes back, and used, not just a starting point that can be bypassed after doing it a couple of time. Instead of pointing out to what exists, basically what a search engine does, it "helps" the user by building. It also gives an amazing sense of agency and power, you "do" get something that seems to come out of nothing, conveniently removing provenance and thus make the user feel quite good about the process.
This is especially poignant to me given this anecdote from a friend I shared just days ago https://news.ycombinator.com/item?id=48457842 showcasing how we wrote a Wacom driver, on his own, without being a developer, thanks to Claude, and how he even potentially helped others by sharing back what he "built" only for someone to suggest an already existing project https://news.ycombinator.com/item?id=48459366 .
nostos/limbus, genoox, engenome, congenica are a few companies/products that I have heard about and have been around for years (the last one was defunct from what I heard last however).
Disclaimer: not affiliated to any of these.
"Like all of the others, this WGS came back non-diagnostic. Unlike the others, my heartbreak had passed the point of hopelessness and I was ready to do something. I contacted every lab we had worked with and requested access to all genomics files. I was going to figure this out myself.
"After a few days, my initial results shocked me. The prototype I built not only accomplished my original goal of confirming that Warren seemed healthy (spoiler: everything is fine), but it found the genetic mutation that took our first son Owen’s life. How could something I built so quickly outperform the top sequencing lab in the country?"
None of us really know what to do!
I was a “such is life” type of person when experiencing tragedy, and fairly ho-hum during joyous moments. With kids, something changed, or “rewired”. I tear up at Bluey episodes and lose sleep due to irrational fears of an early death and not witnessing moments in my kids’ lives.
Having and raising kids is a trip.
In past year I've watched my older (8) start competing on a cheerleading team. Immense tear-streaming joy watching her light up in front of a crowd and build confidence. I was immediately overcome the first time and always feel a strong swell of emotion.
Instead, I have buddies. I wouldn't call it unhealthy, just much more invested than I expected to be.
In case you find it useful, I wrote what I learned in my first year of fatherhood (the little guy is now 17 months): https://alejo.ch/3hj
As you can see, I never found anything as scary as the possibility, however remote, of my baby being unhealthy. So, yeah, best of luck!
Just a joke. It will be marvelous. Savoir the first years. It's a lot of work, bad sleeping nights, but you'll have fond memories for life.
I mean, make no mistake, it is terrifying, but you'll make it through eventually and it will be okay.
Also not sure if this was covered in class, but in case it wasn't: You need to rotate the baby. The baby needs rotation so that the head doesn't grow misshapen. Rotate the baby.
All this to say, all will be good – don't let the rare anecdotes get to you.
[0] https://www.cdc.gov/birth-defects/data-research/facts-stats/...
> It was clear that something about my approach was interesting.
But no approach. Not even a hint.
I do hope it pans out. I do understand it must be a trade secret in order for you to have a business, but I'm still a little underwhelmed.
https://www.janestreet.com/tech-talks/algorithm-for-precisio...
It’s a fascinating story similar to this one, but using “old” AI. The dude spent years learning genetics, and built a program called “medikanren” with Will Byrd. It’s p much a huge graph database of facts they extracted from academic papers, and query using logic programming. It’s a great story
Is this more than a harness built on top of a SOTA commercial LLM?
We're talking about a deep human experience here. I don't know about HN as a whole but I personally come out of this much more touched about the human side of the story and how someone's life events can tremendously change their paths and goals than the LLM itself.
Are you saying that, as an 'AI engineer', you were unaware that LLMs could be used to interpret genetic variants? A thing that Google has been publishing on for well over 10 years?
I'm planning on getting one out in the next few weeks characterizing the system and how it performed on real clinical use-cases vs. alternatives and existing tools.
The TL;DR is that Gamow Labs is a harness and interface company on top of SOTA LLMs as you suggested, but my harness and interface outperforms the existing thing. While this approach would have earned me the "wrapper company" label last year, I hope the success of OpenEvidence, Harvey, Perplexity, and so on has opened minds with respect to the value here.
It was only working through clinical cases that I realized how much more I needed beyond dropping raw reads into Codex.
Your assumption is correct about my technique. I cloned (and expanded) this workflow into an LLM harness, so the LLM is basically orchestrating a bunch of tools that normally humans would use (and writing the conclusions and doing all the standard LLM stuff).
No matter how short his life, or how much his suffering, Owen did get to experience being held by parents who loved him.
Godspeed.
Peering through the NICU window & wondering if my child would survive was one of the most traumatizing moments of my life. Rooting for your work.
oh boy
That said, it’s notoriously difficult to call these types of causal microdeletions, outside of a few known ones. Any insights you make on those efforts would advance the field quite a bit.
As an aside, I have not found SF to be anti-natal but that's because of the community we've formed. Of our friends in SF, almost all are trying for children or have them. Our shared Slack group is full of happy news. Inevitably, many of us must move elsewhere in order to allow them some freedom[2] and good education[3]. So there's a bit of a dead-sea effect, true, but even within that sea there are pockets of community one can find.
0: https://wiki.roshangeorge.dev/w/IVF but also see https://viz.roshangeorge.dev/roshan-genvue/ to see an old set of results and download my VCF here https://my.pgp-hms.org/profile/hu81A8CC and put it through a modern (but not too modern LLM) to get some fun results if you want a starter data set to peek through. Opus can get you far.
1: Fable as expected yields:
Fable 5's safety measures flagged this message for cybersecurity or biology topics. They may flag safe, normal
content as well. These measures let us bring you Mythos-level capability in other areas sooner, and we're working
to refine them. Switched to Opus 4.8. Send feedback with /feedback or learn more
3: Since school system design is determined by both parents and non-parents, it is a vehicle of expression of non-education-related action by the latter, and in a city where they dominate the former, the effects are typical
If you are open to chatting about your experience, I'd love to hear from you. I spend a lot of time learning from and supporting other rare disease families these days.
And the anti-natal thing was kind-of joking not joking. I do know lots of people with kids there now, but when my wife first got pregnant, we were alone.
Does it really outperform it, or you just got an answer faster? Because that's the trap of LLMs.
If they can save more lives without harming other lives I would gladly take it during the analysis. Even saving 1% more lives is an amazing achievement!
It sounds like you just take issue with abortion. That’s more of a you problem than a problem with quantifying baby deaths.
I have no problem with abortion itself.
I'm in the process of getting my first batch of long reads, but I am skeptical that this is the "just" what's needed. There is little doubt that long read > short read, but I think that computational techniques for both need to be improved significantly.
There is already some clinical evidence to support my hypothesis. The first clinical long read trial at Kansas City Mercy showed a 10% bump in diagnostic rate, which is great but not fully solving the problem: https://news.childrensmercy.org/unlocking-answers-faster-chi...
Most Down Syndrome people have happy lives, some can even live alone and have an independent daily live.
Life expectancy is up to 60 years.
Yet in Iceland „Democratization“ of genetic diagnosis lead to basically 0% Down Syndrome kids (births).
Where does this stop? What with someone of a genetic indication of aggressive cancer- life expectancy 55? Abort?
The same (detectable) genetic mutation leads to vastly different lives. This is correlation. Nobody sees the many undiagnosed broken genomes with no visible symptoms that would be aborted with more such businesses like this.
I have two kids. If I could have chosen whether or not they would be born with Down syndrome I would choose would “without” every single time. If either of them were born with it I would love them the same but I would never choose it, if I had the choice to make.
Increasingly though, parents can check, and they are doing, and that means they're getting to pick whether or not to carry a not-quite-perfect baby to term. Many are choosing to terminate and try again. Right now it's for obvious things like Downs, but the scope for what parents choose to terminate can, and probably will, escalate to other detectable problems.
The question is where that ends. Terminating due to a susceptibility to aggressive cancer? Maybe. Due to lower intelligence? Perhaps. Lower physical strength? Probability of autism? Unsymmetrical facial features (e.g. 'ugliness')?
Why/how a pregnancy started (consensually or not for example) or the motivation behind making that choice (just don't want a kid, some genetic diagnosis, etc) are irrelevant. If abortion is bad, it shouldn't be done at all. If it's not bad the reasons behind it should not matter
Always breaks my brain when people argue abortions in case of rape or incest are OK but otherwise a woman should be forced to carry a pregnancy to term, as if it matters to the potential child in the end. Either you're killing a person or not. I'm pro-choice, so it's my opinion that you're not, but the wishy-washy middle ground position is untenable to me.
If that is actually possible someday, legitimately, I am switching to pro-life.
With only rape/incest or mother's certain death as reasons for termination.
I realise this is a weird and nuanced take, but I was adopted and it was back when they forced the mother to keep the kid for 6 months before they'd place tbe kid. Puritanical nonsense?
Make it so accidental pregnancy is 0.0001% (a thousand times less likely than perfect condom and pill/IUD use;) and the above "can fix", and what reason would there be to terminate?
I am anti-death penalty too.
Having conditions on abortion are completely nonsensical and irrational. I don't get it
Is aborting because of an LLM „likelihood“ of some genetic issue (that might lead or not lead to an short or unhappy life) the same as aborting after rape for you?
Biggest problem in fact is that people take genetic markers as deterministic and assume the worst outcome. Lots of false abortions happen. That does not happen with rape.
Yes. Because if you think it's wrong to end a pregnancy, what does it matter how it started or why you want to end it? The rationale behind being "pro-life" is that you feel abortion is killing a child. I happen to believe that a non-viable fetus is not a child, so I am supportive of a woman's right to manage her reproductive system as she so chooses, and therefore it doesn't matter why a woman wants to end a pregnancy--from rape or incest, doesn't want to have a child right now, doesn't want a child with a genetic deformity, or doesn't want a child with blonde hair--it doesn't matter.
On the other hand, if you felt that it was "murder" why would there be any reason good enough to justify murder? The only justifiable reason to kill another human being is in self-defense. A woman who was raped won't be "un-raped" because she is no longer carrying the baby. That sounds harsh, but of course I think it is BS because the whole anti-abortion argument is largely BS.
Access to the tech is probably unequal if it's done privately, which leads to polarization of society where rich people get even more opportunities than poor people. If you want equality of opportunity and an approximately meritocratic society then building a system to prejudice outcomes before kids are even born isn't ideal (although money and education already does this to an extent, those can be countered a bit by government policy; literally growing humans with genetic advantages can't.)
There's a world of potential for choosing foetuses based on criteria that are ethically catastrophic (no girls, no people who are 'impure', etc). You can argue that it's still parental choice even if the parents are terrible people, but normalizing the tech could be a disaster if a future fascist government gets into power. Imagine if the choice was removed from the parents and taken over by the state.
The foetus doesn't get a choice. This is straying very close to anti-abortion rhetoric admittedly, but if you believe that people should get a say in the outcome of their life, then aborting pregnancies based on a possible outcome that might not manifest for decades is very questionable. A baby that gets terminated because current medicine can't stop an aggressive cancer is having the opportunity to wait for medicine to improve taken away from them. Even ignoring the abortion side of things, you can question whether it's right to make that decision on their behalf.
Character, beauty, love, sacrifice. Every one of these involves pain and it makes life worth living. You can't avoid pain, so you might as well engage it in service of those you hold dear.
You should very much care about society wielding a sword like this, because historically we do not wield it well.
It's easy to hand wave the question of pain away, but much wiser men than you and I have arrived at very different conclusions than you suggest.
It's very easy to demand or promote sacrifices you expect other people to make. But I don't find that to be very empathetic.
I've already seen how society is when we shame and hand-wring about the personal decisions others make, and it's not one I want for my kids. At some point you need to be satisfied with your own decisions and then let other people make theirs.
I can be both satisfied with my decisions and still wish better for my neighbor; these are not mutually exclusive.
I am not saying, don't choose between embryos. I'm saying, be careful because it's a slippery slope and not a slide you want to ride.
Eliminating Down syndrome and cystic fibrosis, for example, seems not only reasonable but a moral imperative.
My friend from elementary school has had a Down kid at 19. She takes it well, but managing a grown-up man five times as strong as her (she is petite) when he decides to do something like "take his bicycle and ride on a busy highway" is exhausting.
This is why most people abort Down fetuses.
This is also a hidden cost of having a Down's baby - given how exhausting the care is, other kids either get less attention, or don't get born at all.
Also, the father skedaddled. Allegedly this happens quite often with unhealthy kids...
Answer I heard is hushed "Yes".
I'm not in favor of eugenics but you started from there and now you are sounding "pro-life" each time more as you write more messages in the thread.
Screening an embryo that could not survive outside the womb and deciding to stop the pregnancy based on the results is not the same as "wishing the kid wasn't born".
Your depiction of a person with downs is not representative of the majority.
It gets worse. People with Down’s syndrome face all kinds of health issues. *Half* are born with heart defects. They have higher risk of hearing loss, vision problems, sleep apnoea, thyroid disease, coeliac disease, digestive problems, epilepsy, infections and leukaemia. They also have elevated risk of dementia later in life.
This is a hellish existence which no one would wish on their worst enemy. It's hell for the sufferers and it's hell for the family.
I can’t say for 100% certain but I struggle to believe what you’re describing is more common than not. That strikes me as a level of support provided by the state that most people with ds - and their family - do not enjoy throughout the world.
This is absolutely common in New Zealand.
I think treating people with dignity, respect and part of the community is as important as state financial support.
Our government will also subsidise wages in certain industries for people with certain disabilities. So it’s common for people with intellectual disabilities to be working at supermarkets.
Remember the original context here: someone talked about what they experienced and then you came in questioning it saying how wonderful the support you see is, then said what they experienced is not the majority experience, when the reality is quite the opposite actually. The support you are seeing is what is actually uncommon worldwide. New Zealand is exceptional here.
That said, prospective parents usually don't make their decisions to abort or not to abort on an expected median outcome, but some of the more pessimistic-but-realistic scenarios.
It may be for a different country, both experiences are valid to hear. In my country as far as I know it doesn't happen like yours.
As for trisomy-21, if we wish to increase the rates in the future we can stop performing pre-implantation testing and the MSAFP and nuchal translucency and so on. After all, even if we abort all embryos that carry the markers today, new such embryos will be formed in the future since it comes from non-disjunction so selection pressure once removed will allow it to return.
Potential people aren't actual people though either.
You destroy the life of the family and greatly make the lives of healthy children harder.
Why would you sacrifice your healthy children for one who has no real future?
Complete inversion of common sense.
I.e. not a net-productive member of society? Fun fact, machines are taking everything over, so you and any "healthy children" may be in that same camp very soon. IMO, anyone capable of living happily while being outcompeted by machines is well-adapted for the future.
I'm not; average Down's Syndrome folks seem to be.
> I'm not
Why not?
Oh, the horror.
Listen, you have no idea what these people went through and judging them for not wanting to go through the same thing is shitty.
I found this to be the most interesting theory in the post because I'm not convinced it's at all true.
(And then there was the whole “antinatalist” bit and I am trying really hard to keep this dude’s shoes tied so I will stop there)
Speaking from my own first-hand experience, saying "goodbye" is never easy regardless of the circumstances. Every goodbye has its own texture and none is "easier" than another. They're just...different.
Internet reply: this is extremely disturbing
Its a sad story. I wish them well, I hope this is one of those scenarios that AI works exactly as desired.
This is the sort of performance/quality gap that is not obviously seen (who in the general population understands lab protocols in detail?) and could be massively improved with AI.