HN user
The Smith chart is the electrical engineer's favorite: https://en.wikipedia.org/wiki/Smith_chart
You either love it or hate it, depending on how well your electromagnetics class was taught.
Unpaywalled: https://archive.is/dmp0m
The cynical part of me wonders: if this has been a promising approach for 10+ years, why weren't they able to secure VC funding years ago (or nonprofit biomedical research funding from places like the Gates Foundation that care a lot about infectious disease)?
Thanks for the response! This is very cool and sounds like a reasonable plan. Best of luck!
I am totally onboard with the premise (as a TechBio-adjacent person), and some of the approaches you're taking (focused domain-specific models like Orthrus, rather than massive foundation models like Evo2).
I'm curious about what your strategy is for data collection to fuel improved algorithmic design. Are you building out experimental capacity to generate datasets in house, or is that largely farmed out to partners?
There are biotech companies like Eikon Therapeutics (https://www.eikontx.com/ ) where super-resolution microscopy in living cells is a central part of the platform.
There is also one widespread approach that isn't mentioned in the article: expansion microscopy. Expansion takes the scifi-sounding approach of: what if you could make your sample physically bigger? See the Wikipedia page for more: https://en.wikipedia.org/wiki/Expansion_microscopy
A thread from yesterday about why gene therapy hasn't reached its potential: https://news.ycombinator.com/item?id=44573193
DeepMind/Google does a lot more than the other places that most HN readers would think about first (Amazon, Meta, etc). But there is a lot of excellent work with equal ambition and scale happening in pharma and biotech, that is less visible to the average HN reader. There is also excellent work happening in academic science as well (frequently as a collaboration with industry for compute). NVIDIA partners with whoever they can to get you committed to their tech stack.
For instance, Evo2 by the Arc Institute is a DNA Foundation Model that can do some really remarkable things to understand/interpret/design DNA sequences, and there are now multiple open weight models for working with biomolecules at a structural level that are equivalent to AlphaFold 3.
Very cool. There are also methods that allow you to extract some notion of motion from variability in CryoEM data, e.g. CryoDRGN-ET [1].
I'm curious if you've worked with any of those models and how they relate to NMR data and MD simulations.
thought this was about tracing neural circuits in the brain and was disappointed.
Detailed New England Journal of Medicine article about this case: https://www.nejm.org/doi/full/10.1056/NEJMoa2504747
And an Editorial piece (more technical than the NYT): https://www.nejm.org/doi/full/10.1056/NEJMe2505721
It is possible that they are licensing technology that was developed in academic science and are raising money to scale it up and get it ultra-standarized for commercial scale.
I agree that the modern Silicon Valley model of VC funding has been spoiled by SaaS startups, where the capital expense is smaller, the timeline to exit is shorter, and pivots are easier. It is not great for deeptech innovation because those require more capital, time, and are more technology-constrained than software. Ironically, modern VC was developed to support semiconductor startups (1970s-90s), but has drifted from that technology-heavy origin.
Indeed, now is the moment to step on the gas in biotech. The past 15 years have been nothing short of extraordinary in the field. We finally have the tools needed to effectively measure biology, manipulate biology, and increasingly predict biology. More recently, we have been able to turn more and more problems into computational problems.
With all of this coming together, we should be accelerating both public and private investment in biotechnology because we're getting closer and closer to transformative therapies. But...we're failing to rise to the occasion and meet the moment.
It is worth mentioning that China is heavily investing in biotechnology and they are getting genuinely good at the more commodified parts of the industry. This blog post [1] is long and aimed at a biotech expert audience, but one summary line that stands out is that "the drug industry is having its own DeepSeek Moment" [2].
To that end, I believe that this is the time to invest in the US biotechnology ecosystem so that we remain competitive with China. The ongoing crisis at the NIH is antithetical to this goal, as Derek Lowe's blog posts describe.
[1] https://centuryofbio.com/p/commoditization [2] https://www.wsj.com/health/pharma/the-drug-industry-is-havin...
Lots of biotech companies are working on immunosuppressant-free islet-equivalent transplantation.
Two examples off the top of my head: Sana recently announced islet cell transplantation without immunosuppression (press release: https://ir.sana.com/news-releases/news-release-details/sana-... ) and Vertex (ongoing trial: https://www.breakthrought1d.org/news-and-updates/vertex-laun... ).
1) Creative projects beget creative projects. When you start working on one project, you'll have ideas for a dozen more, and probably one of those might actually be a good project idea to continue exploring and refining and shaping into a reasonable problem. It's really hard to come up with things if you're just staring at a blank sheet of paper, but working on anything at all gets this virtuous cycle started.
2) Talk to people! Bounce your ideas/areas of excitement off of other people, and see what gets reflected back at you. That signal can be very helpful to see when you've stumbled across an idea or problem that might be useful to more people than just yourself, i.e. a more important area of investigation.
3) Read, read, read. And take notes on random ideas you have while reading, and things that papers missed or didn't look into. If you do this enough and take some time to reflect on it, you can start to find gaps in knowledge that could be addressed.
I am pretty sure that the CGP Grey video was the core inspiration for this article. Nearly every image in the article was featured in the YouTube video, which predates the article by ~4 years.
Have you looked into Nebula Genomics? They will do 30X coverage WGS for $299, and will give you the underlying data so you can analyze it on your own (in addition to their analyses).
Nebula Genomics: https://nebula.org/whole-genome-sequencing/
I would recommend the edX Introduction to Biology course [0].
It is a simplified version of the introductory biology course at MIT, that doesn't _focus_ on naming/defining things in biology. Instead, it uses the lenses of genetics and biochemistry to explore how the core machinery of life works, and how we got to our current understanding. That said, there is some amount of memorization that is unavoidable.
[0] https://www.edx.org/course/introduction-to-biology-the-secre...
The Vital Question, by Nick Lane is a pretty good book that takes a thermodynamics first approach to this question.
Mostly agreed. In my eyes, the takeaway from this piece is not that neuroscience is futile and that we should give up, but rather that we now should redouble efforts on analysis methods to make sense of the large, complex datasets that we are on the cusp of generating.
The stories in this piece really highlight the importance of having these tough conversations with family members about what you want to happen for your own medical care if the worst happened.
I really enjoyed Being Mortal by Atul Gawande (author of The Checklist Manifesto), that tells intensely personal stories about, well, the process of dying, and the increasingly prolonged tug of war between medicine and death.
One thing that may be a bit of a challenge is how quickly things change in the technology world. "Code Status" is medical lingo for the descriptor of what the patient expresses they want to have happen if their heart or breathing were to stop. Most people are full code - CPR, mechanical ventilation, etc. But patients can choose to be DNR/DNI, meaning "Do Not Resuscitate, Do Not Intubate", meaning very limited interventions would be performed.
As the tech gets better, I wonder if a more sophisticated decision tree might be needed in the future -- if XYZ happens where 30% of patients make a recovery, begin ECMO, but if ABC happens in which only 5% of patients recover, do not start ECMO.
On the topic of describing the events of the code (cardiopulmonary resuscitation) in great detail, there have been studies showing that bringing the family into the room during a code leads to decreased PTSD for the family.
From a study in the New England Journal of Medicine: "Conclusions: Family presence during CPR was associated with positive results on psychological variables and did not interfere with medical efforts, increase stress in the health care team, or result in medicolegal conflicts."
Full study: https://www.nejm.org/doi/full/10.1056/NEJMoa1203366
DOI for Sci-Hub: DOI: 10.1056/NEJMoa1203366
The Allen Institutes are incredible organizations. They do these enormous Big Science projects that no single lab could ever do, generating beautiful datasets that the entire scientific community can use freely. All of their data is publicly available. Paul Allen's contributions to accelerating science are immense and cannot be understated.
Grape juice is the notorious one here. It doesn't have a very strong taste on its own, so when blended basically serves as a not-technically-added-sugar source of added sugar.
For those who might not know, the University of Edinburgh had a department of artificial intelligence in the 1970s. They were very forward thinking at the time -- it later got folded into the School of Informatics, but Edinburgh remains one of the best places to work on AI/ML work.
edit: slightly awkward phrasing in my original comment above. Amended: They were (and still are!) very forward thinking.
Different kind of tech. Boston is the world's capital of biotech.
Just to add a single word -- I think it is very important that investments be well-targeted. And almost all of the time, this means listening to the scientists involved to get a realistic picture of where the domain experts think the field is going and how to best support efforts that are likely to be successful.
Sometimes, we have some promising leads on disease that we can begin to chase down. Most of the time, however, we have no idea where to begin, and so a shotgun basic science approach may ultimately be more fruitful. Every subfield is different and difficult to master, which is why it is so important to engage with the people who actually do the work.
I know the tone of this sounds like I'm beating a dead horse. I'm sorry.
Most recent edition was published in 2012. It's extensively updated between editions, so very much worth reading if you are seriously interested in neuroscience.