It’s interesting how Egypt’s efforts to monitor and test for malaria contributed to this accomplishment. It underscores how eradicating many infectious diseases will require a deep understanding not only of the disease itself, but also the cycles of transmission and the complex ecology of different hosts.
Malaria’s complex lifecycle [1] seems like it would be easy to “break” with different interventions, but we’ve seen historically malaria has been difficult to eradicate. Why is this?
I think the greatest challenge with eradicating Malaria is that it is most prevalent in impoverished regions of the world. The USA occasionally has incursions of Malaria which is quickly quashed by the CDC National Malaria Surveillance System. If you have enough funding, Malaria is preventable. However, if most people do not have access to medical care, they cannot be diagnosed or tracked.
Essentially, a lack of access to health care results in Malaria continuing to devastate regions of the world. If you ever want to save a life, donating to the MSF is a great way to do it.
>If you have enough funding, Malaria is preventable.
It requires more than funding to solve the problem. Sorry that my source is a YouTube video, but https://www.youtube.com/watch?v=CGRtyxEpoGg explains a general problem (that of trying to solve problems that are more prevalent elsewhere in the world, from within your own cultural context) and gives malaria as an example. People in malaria-afflicted countries, given free insecticide-treated nets, will often try to use them for fishing - not caring about the effect the insecticide will have on the haul. It's not due to ignorance or a lack of understanding, but due to a value judgment: people who have lived with malaria for generations don't see it as being as big of a problem, while poor people (on a global scale - not like in the US where "the poor" can afford some really impressive things) are always concerned with food supply.
So, once you understand the cultural context, you know to give the people both mosquito netting and fishing nets. This would be easier if you could afford outreach and food assistance. This just shows that food assistance is a key part of medical assistance.
> people who have lived with malaria for generations don't see it as being as big of a problem
I don't accept the idea that these people want to live with malaria because it is normal. People don't like being bitten by insects. They just like starving to death much less. Appropriate funding can honestly solve this problem.
>So, once you understand the cultural context, you know to give the people both mosquito netting and fishing nets.
Another example of this problem was the distribution of high efficiency stoves as a form of carbon credits. People just used both their low efficiency stove and the higher efficiency stove to increase yield.[0] If you give someone who needs more nets a fishing net and a mosquito net, guess what they're going to do. This is a fundamental methodological issue, not a simple problem of "Okay, but now we understand."
>I don't accept the idea that these people want to live with malaria because it is normal.
Of course not, but people are also capable of making their own decisions about what is affecting their lives most immediately. We just saw a massive number of educated populations in the US refuse vaccination efforts during a global pandemic because of a risk tradeoff, despite that decision statistically making no sense for the overwhelming majority of them. You think someone impoverished and facing food scarcity is going to prioritize a government or NGO effort to solve a problem that is inherently a low statistical background noise to their life experience? Why would they?
> If you give someone who needs more nets a fishing net and a mosquito net, guess what they're going to do.
So you give them a mosquito net and two fishing nets. Or five. Put a giant mosquito label on the mosquito net and a giant fish label on the fishing nets.
> This is a fundamental methodological issue, not a simple problem of "Okay, but now we understand."
Like.. way to overcomplicate something that is indeed solved with more money.
If you are giving other forms of aid, you can incentivize the use of netting. However, netting is just the beginning of such an effort. For instance, you don't see people in Louisiana sleeping with mosquito nets. This is because there are eradication efforts using pesticides and sterilized males. More money means that you can engage in that as well as helping locals avoid bites.
This. That malaria is not prevalent in the Southern U.S. (there's a reason the CDC is in Atlanta) is as much an economic choice as an epidemiological success story.
"Carpet bombing" is perhaps a hyperbolic term, but widespread application of DDT in the southeastern US was, in fact, a central component of the effort.
It's very possible getting rid of malaria made this was a worthwhile, even given our modern knowledge, given the treatment options available at the time.
Large-scale medical treatments are always a difficult area, because almost no treatment, or course of action, is risk-free, but malaria was awful when it was more widespread.
It's such a neat method because it is so inexpensive too. You take a bunch of mosquitos, irradiate them with just the right dose at the right time and then release them en-masse.
Its not really that we know better. We known more and we know there is more of a trade of than was assumed then.
But to know better would mean we would have done anything different back then. If the choice is a silent spring (hyperpole, but okay) or dead babies from malaria in the US, no politician is going to align with the "I support dead babies party" and nobody is going to listen to those who do.
Until they banned DDT ostensibly because it was a threat to 'wild life'. I'm sure it affected people very adversely (it's rumored that DDT was one of the major causes of polio). Right now there will be chemicals which are widely used which fall in the same league.
If I may add two links for people interested in helping people in impoverished regions.
It seems that people on the ground living there also really need basic things like mosquito nets, clean drinking water, proper nutrition, medical equipment, facilities where they can be treated, medicaments, and so on.
> If you have enough funding, Malaria is preventable
Malaria is also dependent on a non-human vector. That means you can target it without requiring peoples' co-operation. Contrast that with e.g. polio where you have to convince people to get vaccinated.
I agree, once you detect that malaria is present, you can interrupt the cycle by treating wetlands in the area to kill mosquitos. These efforts are crazy expensive, but then again, so are the health care costs related to treating the disease. In the USA they use a ton of different methods to accomplish this, with and without pesticides. But it's really expensive.
Malaria has multiple dependencies but they’re all resilient like well set up k8s. You can reduce its function by attacking multiple paths but, mathematically, to destroy it one of the decencies has to go to 0 or several have to be severely degraded. Polio was comparatively easy because it had a cheap vaccine you could take by mouth and you could isolate
For those like myself who design proteins for a living, the open secret is that well before AlphaFold, it was pretty much possible to get a good-enough structure of any particular protein you really cared about (from say 2005) by other means, namely Baker’s Rosetta.
I constantly use AlphaFold structures today [1]. And AlphaFold is fantastic. But it only replaces one small step in solving any real-world problem involving proteins such as designing a safe, therapeutic protein binder to interrupt cancer-associated protein-protein interactions or designing an enzyme to degrade PFAS.
I think the primary achievement is that it gets protein structures in front of a lot more smart eyes, and for a lot more proteins. For “everyone else” who never needed to master computational protein structure prediction workflows before, they now have easy access to the rich, function-determinative structural information they need to understand and solve their problem.
The real tough problem in protein design is how to use these structure predictions to understand and ultimately create proteins we care about.
Forget Rosetta. Even installing that shit was hard, and running it on a sufficiently beefy machine was probably really not a thing in the late aughts. For protein design you mostly just need a quick and dirty estimate of what it looks like, and you have friend proteins that can be used to homology map, you could just use phyre/phyre2, which is an online threading model and be close enough to get work done. Upload the pdb, upload the sequence, bing bam boom.
Agreed, the UX of a magnet timer on the fridge beats using any kind of smart device for the task of setting kitchen timers. Most of them start a simple count up if not programmed for a specific duration, so you can watch the seconds.
Most magnet timers also remember the last duration, so if you use the same timer a lot for the same task (tea, for example), it’s literally a single button press. The same operation on any kind of smart device contains a staggering number of steps, each of which requires cognition and attention.
Magnet timers are also super cheap, so you can get another one if you have two favorite durations. A simple solution meets a simple problem
Agreed on the hopes that these methods lead to novel biocatalysts (but they aren’t quite there yet).
David Baker’s lab has recently published on using their own diffusion model (RFdiffusion) to design novel biocatalysts that perform hydrolysis using a catalytic triad of serine, aspartic acid, and histidine, as well as an oxyanion hole, which is much more complex than the binders designed by AlphaProteo [1].
It gives me hope that we’ll soon be able to design biocatalysts as good as natural ones, but for any problem we care about.
Good point! And a related topic, we call the organism that lives happily in our gut E. coli and we also call the organisms that cause disease the same name. What’s the difference?
It turns out that the Escherichia coli (to spell out its Latin binomial) that cause disease are in some sense “diseased” themselves: the genes that enable them to be pathogenic, or make them pathogenic, I should say, are originally from a phage, a type of virus that infects bacteria [1]. In a manner that is not the same as, but conceptually similar to how HIV inserts its genes into the human’s genome, phages insert their genes (termed the “prophage”) into the bacterial genome.
In addition, most strains of pathogenic Escherichia are also holding on to an entirely separate, small, circular “genome” called a plasmid, also of exogenous origin, that contains additional genes that make them pathogenic.
So in addition to wide genome variation within the “species” (which is not really the same thing for bacteria as for mammals, mind you) of Escherichia coli, many subtypes have additional genetic material from endogenous sources that substantially changes their observed characteristics (phenotype).
Also happen to be in microbiology, but pretty far from the medical side of things.
Do you have a citation on the fact that 'most' pathogenic strains have a plasmid making them so? Some guys in our lab have been playing around with plasmid copy number lately (in a largely 'basic science' kind of way) -- this could give some nice context for that work.
The flip side of this is that progress in ML for biology is always going to be _slower_ than progress in ML for natural languages and images [1].
Humans are natural machines capable of sensing and verifying the correctness of a piece of text or an image in milliseconds. So if you have a model that generates text or images, it’s trivial to see if they’re any good. Whereas for biology, the time to validate a model’s output is measured more in weeks. If you generate a new backbone with RFDiffusion, and then generate some protein sequences with LigandMPNN, and then want to see if they fold correctly … that takes a week. Every time. Use ML to solve _that_ problem and you’ll be rich.
TFA mentions the difficulty of performing biological assays at scale, and there are numerous other challenges. Such as the number of different kinds of assays required to get the multimodal data needed to train the latest models like ESM-3 (which is multimodal, in this context meaning primary sequence, secondary structure, tertiary structure, as well as several other tracks). You can’t just scale a fluorescent product plate reader assay to get the data you need. We need sequencing tech, functional assays, protein-protein interaction assays, X-ray crystallography, and a dozen others, all at scale.
What I’d love to see companies like A-Alpha and Gordian and others do is see if they can use the ML to improve the wet lab tech. Make the assays better, faster, cheaper with ML. Like how they use ML to translate the electrical signals of DNA passing through the pore into a sequence in the Nanopore sequencers. So many companies have these sweet assays that are very good. In my opinion, if we want transformative progress in biology, we should spend less time fitting the same data with different models, and spend more time improving and scaling wet lab assays using ML. Can we use ML to make the assay better, make our processes better, to improve the amount and quality of data we generate? The thesis of TFA (and experience) suggests that using the data will be the easy part
Probably worth mentioning that David Baker’s lab released a similar model (predicts protein structure along with bound DNA and ligands), just a couple of months ago, and it is open source [1].
It’s also worth remembering that it was David Baker who originally came up with the idea of extending AlphaFold from predicting just proteins to predicting ligands as well [2].
Unlike AlphaFold 3, which predicts only a small, preselected subset of ligands, RosettaFold All Atom predicts a much wider range of small molecules. While I am certain that neither network is up to the task of designing an enzyme, these are exciting steps.
One of the more exciting aspects of the RosettaFold paper is that they train the model for predicting structures, but then also use the structure predicting model as the denoising model in a diffusion process, enabling them to actually design new functional proteins. Presumably, DeepMind is working on this problem as well.
I appreciated this, but it's probably worth mentioning: when you say AlphaFold 3, you're talking about AlphaFold 2.
TFA announces AlphaFold 3.
Post: "Unlike AlphaFold 3, which predicts only a small, preselected subset of ligands, RosettaFold All Atom predicts a much wider range of small molecules"
TFA: "AlphaFold 3...*models large biomolecules such as proteins, DNA and RNA*, as well as small molecules, also known as ligands"
Post: "they also use the structure predicting model as the denoising model in a diffusion process...Presumably, DeepMind is working on this problem as well."
TFA: "AlphaFold 3 assembles its predictions using a diffusion network, akin to those found in AI image generators."
Coming up with ideas is cheaper than executing the ideas. Predicting a wide range of molecules okay-ish is cheaper than predicting a small range of molecules very well.
As cool as this is, the word “microplastics” is a little misleading. There are dozens of types of plastic in common use, each made from a different monomer with a different chemical linkage, of which PET is only one. The engineered protein in TFA will only work on PET and we’ll need to design new proteins for the other types of plastic. (I can help with that.)
The problem with enzymes eating plastic is that enzymes are small Pacman-shaped protein blobs that are maybe 10 nanometers in diameter, whereas things made of plastic like bottles or even microplastics are huge in comparison. How do you get the little Pacman jaws around the bottle to start breaking it down?
The research paper [1] describes the authors’ effective innovation. They make a protein where one end is a pore-forming shape, and the other end is a PET cutting (called a PETase in the jargon of the field). This way, their protein can access nooks and crannies in the macroplastic shapes, allowing tons of copies of this small enzyme to fully degrade a bottle.
Without this, a great deal of physical agitation is required to break down the plastics into small enough chunks that earlier Pacman enzymes could work on, increasing the time and the cost.
I hope we’ll see the idea of linking the enzymatic “scissors” to a protein pore be used to engineer enzymes to degrade other types of plastics in the future, as the general idea of getting the catalytic machinery into physical contact with every bit of the bottle is broadly applicable to all plastics, not just PET (which is great news)
I don't really see why there is a problem with degrading whole bottles. If you have separated from the waste stream, you can incinerate or even landfill them (it's not like you'd be wasting any resources). It's the microplastics that form when the bottles are dumped into the oceans or waterways and broken up by Nature which need a novel solution for removal.
I think what lysozyme said holds even if you do not consider "whole" bottles at all. According to wikipedia the biggest microplastics are 5mm, whereas the enzyme is 10nm, that is 5 orders of magnitude of difference, but just in lenght! To process the entire volume you need to cube the units and you get 15 orders of magnitude in difference (1 mm^3 to 10 nm^3).
To get an idea I asked wolfram alpha what is the volume of the average human, and apparently that is around 66 liters. Then I looked up the estimated water volume of the Baltic sea, and wikipedia says it is 21,700 km^3 of water, soo
$ units
586 units, 56 prefixes
You have: 21.7E3 km3
You want: 66 liters
* 3.2878788e+14
/ 3.0414747e-15
if you could somehow fill your entire body with water, then make 30 copies of yourself, and you (30 of you) drink an entire Baltic sea (one for each), that is a very very rough analogy of the task we are giving to that poor enzyme. And this is for a single speck of microplastic! Of course the enzyme is not alone, there are a few other billions (trillions?) others with it, but there are also a few million specks of microplastic at any point in the sea. This is a very difficult task.
Plastic aren't simply plastics, they have lots of additives to give them different properties. Incinerating is transforming these additives into other chemicals maybe making more toxic molecules escape to the environment in ashes, dust, smoke. And landfilling is storing trash for future generations to solve the problem.
The most plausible reason they might need to dig them up is to remediate them. Landfills require maintenance in perpetuity, which costs a considerable amount of money. The biggest expense is maintaining the top cap—if it leaks, big problems can result.
After several centuries, it’s hard to imagine that most landfills will still be doing regular maintenance and fighting off entropy maintaining the cap. At some point, with the right technology, it becomes more sensible to reprocess the waste in a more permanent manner.
I think plasma gasification is likely the best idea, but it still needs work.
I would agree. Burying our post-nuclear family waste worldwide for the last 60 years will come back to haunt us. I also agree that we’ll have the tech to address it then into a more sustainable solution and possibly extract energy from it.
Looks like the title was edited to accommodate your clarification.
Unfortunately, that makes your comment a bit confusing, since the context of the title change is not present. I think the best solution would have been the title, "Scientists create artificial protein capable of degrading [PET] microplastics in bottles".
It was never about precision, truth, nor actual science. It was always about "plastics=bad" ideological virtue signaling, just like "chemicals=bad" and "(non-ionising) radiation=bad" before it.
> It was never about precision, truth, nor actual science. It was always about "plastics=bad" ideological virtue signaling
Microplastics were not a concept created for ideological virtue signaling. I don't know who manipulated you into thinking that was true, but you may wish to re-evaluate where you've been getting your information. The good news is that you don't have to depend on some invented sinister backstory for microplastics, you can instead read the paper where the phrase was coined for yourself (https://www.researchgate.net/publication/8575062_Lost_at_Sea...) and see that it was just a lot of typical boring science like searching through sediment and plankton samples, and keeping track of what lugworms eat. A paper that concludes with "we'd need more research to determine if there are any environmental consequences" is about as far from ideological virtue signaling as it gets. Take your own advice and "Beware those who distort the truth and exploit fear for their own gains."
As far as precision goes, currently microplastics are for plastic bits smaller than 5mm. We even have primary and secondary categories for them. Nanoplastics are for bits smaller than 100 nm. Do we really need a better classification system at this stage? I imagine that shortly after we do, we'll get one. Science loves to come up with boxes to put things in.
Regardless what the origin was, it should be clear to see that the term has become associated with ideological virtue signaling and used to promote all manner of junk science and taken up by the engagement-chasing paranoia-spreading media.
Great article. I think we find time and time again that the best leaders are really “cheerleaders”. It turns out that inspiring people, personally living the values you espouse, and giving your team trust and respect is the key to success.
In my experience, people excel in environments where they are given a high level of trust, autonomy, and clear goals on the timescale of months and years.
Teams on the rack of egotistical middle managers get stretched thin until they break. There aren’t a lot of good examples of “micromanagitis” being cured. So-called leaders taking credit for their team’s work leads to their team not trusting that leadership has their growth in mind. Art Blakey’s real legacy is both the amazing players he mentored and tutored, but also the attitude that if you give people the tools they need, they’ll excel
I did try to do this, intermittently. A couple people are now VP's or CTO's, and one founded an early (successful) Internet company; whether I had anything to do with it, who knows?
I think parent has hit on the how and GP has hit on the why.
How LLMs are able to give convincing wrong answers: they “can predict the correct ‘shape’ of an answer” (parent).
Why LLMs are able to give convincing wrong answers is a little more complicated, but basically it’s because the model is tuned by human feedback. The reinforcement learning from human feedback (RLHF) that is used to tune LLM products like ChatGPT is a system based on human ranking. It’s a matter of getting exactly what you ask for.
If you tune a model by having humans rank the outputs, despite your best efforts to instruct the humans to be dispassionate and select which outputs are most convincing/best/most informative, I think what you’ll get is a bias towards answers humans like. Not every human will know every answer, so sometimes they’ll select one that’s wrong but likable. And that’s what’s used to tune the model.
You might be able to improve this with curated training data (maybe something a little more robust than having graders grade each other). I don’t know if it’s entirely fixable though.
The brilliant thing about the parent’s comment about the “shape” of the answer is that it reveals how much humans have (uh, historically, now, I guess) relied on the shape of information to convey its trustworthiness. Expand the notion of “shape” a bit to include the medium. If somebody bothered to take the time to correctly shape an answer, we take that as a sign of trustworthiness, like how you might trust something written in a carefully-typeset book more than this comment.
Surely no one would take the time to write a whole book on a topic they know nothing about. Implies books are trustworthy. Look at all the effort that went in. Proof of effort. When perfectly-shaped answers in exactly the form you expected are presented in a friendly way and commercial context, they certainly read as trustworthy as Campbell’s soup cans. But LLMs can generate books worth of nonsense in exactly the right shapes without effort, so we as readers can no longer use the shape of an answer to hint at its trustworthiness.
So maybe the answer is just to train on books only, because they are the highest quality source of training data. And carefully select and accredit the tuning data, so the model only knows the truth. It’s a data problem, not a model problem
> The brilliant thing about the parent’s comment about the “shape” of the answer is that it reveals how much humans have (uh, historically, now, I guess) relied on the shape of information to convey its trustworthiness.
This is the basis of Rumor. If you tell a story about someone that is entirely false but sounds like something they're already suspected of or known to do, people will generally believe it without verification since the "shape" of the story fits people's expectations of the subject.
To date I've decried the choice of "hallucination" instead of "lies" for false LLM output, but it now seems clear to me that LLMs are a literal rumor mill.
Malaria’s complex lifecycle [1] seems like it would be easy to “break” with different interventions, but we’ve seen historically malaria has been difficult to eradicate. Why is this?
1. https://en.m.wikipedia.org/wiki/Plasmodium#/media/File%3ALif...