Can AI Crack The Biology Code?

As artificial intelligence seeps into some realms of society, it rushes into others. One area it's making a big difference is protein science — as in the "building blocks of life," proteins! Producer Berly McCoy https://www.npr.org/people/985775371/berly-mccoy talks to host Emily Kwong https://www.npr.org/people/767284140/emily-kwong about the newest advance in protein science: AlphaFold3, an AI program from Google DeepMind. Plus, they talk about the wider field of AI protein science and why researchers hope it will solve a range of problems, from disease to the climate. Have other aspects of AI you want us to cover? Email us at [email protected] mailto:[email protected]. See pcm.adswizz.com https://pcm.adswizz.com for information about our collection and use of personal data for sponsorship and to manage your podcast sponsorship preferences. NPR Privacy Policy https://www.npr.org/about-npr/179878450/privacy-policy

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  This message comes from NPR sponsor, Viori, featuring the Sunday Performance Jogger. Receive 20% off your first purchase at viori.com/npr on any US orders over $75 and free returns. Exclusions apply. Visit the website for full terms and conditions.

• 0:19Emily Kwong?

  You're listening to Short Wave from NPR.

• 0:24Emily Kwong?

  Hey, hey, Shortwavers. Emily Kwong here with producer Burley McCoy. What's up, Burley?

• 0:29Speaker 02

  Hey, Emily.

• 0:29Emily Kwong?

  Hello. What do you have for us today?

• 0:31Speaker 02

  OK, so Emily, today I want to dig into how AI has shaken up the field of protein science, as in the fundamental building blocks of life: proteins.

• 0:41Emily Kwong?

  I've heard of them, yeah. I mean, this is like what you studied back in your scientist days.

• 0:46Speaker 02

  Yes, yes, I love proteins.

• 0:48Emily Kwong?

  Oh, we love that you love them. How has AI moved the needle in this field, though?

• 0:52Speaker 02

  Well, scientists have used it to dig into a problem that protein scientists have struggled with for more than 60 years. And that is, what do these building blocks, of which there are millions, look like?

• 1:04Emily Kwong?

  Like their shape?

• 1:06Speaker 02

  Like their shape, yeah, exactly.

• 1:07Emily Kwong?

  And why is that so important?

• 1:08Speaker 02

  Well, the ability of a protein to do its specific job, so like, carry oxygen through your body or turn light into sugar, that relies wholly on its unique, complicated shape. So to understand how it works, you need to know its shape.

• 1:22Emily Kwong?

  But why can't scientists just run an experiment to determine the shape?

• 1:26Speaker 02

  They can for some proteins, but those experiments can take years and years. And Emily, that's because a scientist essentially needs to take the equivalent of a molecular photo of the protein to map its complicated shape. But getting the protein to cooperate to get that photo, so like to hold still, for example, without falling apart, that can be super tricky. And it could take a grad student's entire PhD program to figure out a single protein. And other proteins were just abandoned because they would never cooperate.

• 1:58Emily Kwong?

  Proteins sound difficult, honestly. So the challenge is, how do you figure out a protein's shape without running these super tedious experiments? Is this where AI comes in?

• 2:11Speaker 02

  Yeah, and to give you a sense of kind of how AI has changed the protein game, there's this protein competition that scientists run every other year.

• 2:20Emily Kwong?

  Get out a protein competition, OK.

• 2:22Speaker 02

  Yeah, and they've run it for the past 30 years, where groups will basically compete on who can accurately guess the most protein shapes. It's like nerd central, for sure.

• 2:32Emily Kwong?

  We love.

• 2:33Speaker 02

  And for most of that 30-year history, participants have really only made incremental progress. But in 2020, Google DeepMind used AlphaFold 2, that's its AI protein prediction model, and Emily. AlphaFold 2 blew the other competition out of the water completely.

• 2:51Emily Kwong?

  Wow, OK.

• 2:52Speaker 02

  Game changer. And now, the Google DeepMind team has taken this AI tool to the next level by expanding it beyond proteins.

• 3:03Emily Kwong?

  So today on the show, how scientists have taken a huge step to understanding the building blocks of life using AI.

• 3:10Speaker 02

  Plus, how other researchers are using the tech to design brand-new proteins, ones never before seen in nature.

• 3:17Emily Kwong?

  And how AI could help us solve the biggest problems we face today, from disease to climate. You are listening to Short Wave, the science podcast from NPR.

• 3:47Speaker 00

  Stay tuned for the following announcements and messages from our sponsors.

• 3:51Speaker 03

  This week on Here and Now Anytime, in the closing days of the midterm elections, we're in Iowa and Arizona, two states where independent voters will decide who controls Congress for the next two years.

• 4:02Speaker 04

  Even though we're in a red district, we will get out the Democratic and independent vote, we will.

• 4:09Speaker 03

  Here and Now Anytime, listen on the NPR app or anywhere you listen to podcasts.

• 4:16Speaker 05

  Hey, it's Ray from Car Talk. Are you tired of all the depth and thoughtful care that goes into NPR shows? Want some good old fashioned goofing around and stumbling to figure out what's going on? Well, I've been taking occasional car questions again. You can hear them by signing up for NPR Plus, along with lots of other bonus content. Just go to plus.npr.org.

• 4:39Speaker 06

  This message comes from ProLon. Want to look and feel your best this fall? ProLon's five-day fasting-mimicking diet is a clinically-developed nutrition program with the goal of promoting fat loss while protecting lean body mass. Developed at USC's Longevity Institute, it aims to assist the body in entering a fasting-like state that helps reset metabolism, target visceral fat, and support healthy metabolic markers. Get 15% off plus a bonus gift when you subscribe at prolonlife.com/npr.

• 5:09Emily Kwong?

  Okay, Burley, so scientists, it seems, have been trying to figure out the complicated shapes of proteins for decades to better understand how they work. Why has this been such a complicated thing to figure out?

• 5:21Speaker 02

  Well, the short answer, Emily, is that there are so many theoretical ways a single protein could fold that it's a big problem to solve. So if you unfolded a protein, it would kind of look like a bunch of beads on a long string. Those beads are little molecules called amino acids.

• 5:38Emily Kwong?

  Oh, I remember this from biology. There are like 20 types of amino acids. Each one is a little different.

• 5:46Speaker 02

  Right, so each one has a slightly different shape, and that kind of dictates how that part of the string can be folded up. Because proteins often have 100 or more amino acids, you can see how imagining all the ways it could fold would get complicated.

• 6:00Emily Kwong?

  Yeah, it just sounds like thousands of different shapes, or what, hundreds of thousands of different shapes?

• 6:06Speaker 02

  Okay, try billions of trillions, Emily. Like, there are theoretically more ways for one single protein to fold than there are stars in our night sky.

• 6:15Emily Kwong?

  This sounds like a glorious nightmare.

• 6:16Speaker 02

  Right?

• 6:17Emily Kwong?

  I'm so curious. Okay, so you said that AI has helped us make some leaps and bounds towards a solution. How does this technology work?

• 6:27Speaker 02

  So this AlphaFold model is a type of AI called a deep learning program, which is this huge network of data processing points called nodes. And the purpose of this network is to learn and then make predictions based on what it's learned. In AlphaFold's case and other models like it, it learns about proteins from a huge collection of protein structures that scientists have been building on for decades from their experimental data.

• 6:53Emily Kwong?

  Okay, so the idea is that after these models use all of that carefully gathered experimental data to learn, they can then predict the shapes of proteins they do not know yet.

• 7:05Speaker 02

  Exactly.

• 7:06Emily Kwong?

  Okay, and going back to the protein competition in 2020, how did AlphaFold blow away the competition?

• 7:12Speaker 02

  So they essentially changed the whole architecture of their model. They had been using AI before, but remember the beads on a string analogy? If amino acids are the beads, even if one bead is far from another on the string, when it all folds up, they could be right next to each other. So with AlphaFold 2, the model looked at distances between all of the different amino acids and previous knowledge from solved protein structures.

• 7:37Emily Kwong?

  Awesome.

• 7:38Speaker 02

  And the accuracy and speed of the predictions went way up.

• 7:42Emily Kwong?

  Okay, I'm assuming that made a huge difference for scientists everywhere studying proteins.

• 7:47Speaker 02

  Totally. Julien Bergeron, a structural biologist at King's College London, is one of them. He studies the tail-like appendage that propels bacteria. So it's called a flagellum, and it's pretty complicated.

• 8:01Speaker 07

  It's this huge assembly. So it's longer than the bacterial cell itself. It consists of 20 to 25 different proteins, but many of them have hundreds of thousands of copies of that protein.

• 8:14Speaker 02

  And these huge propeller machines are what gives some bacteria the ability to make you sick or build plaque on your teeth. So Julien's lab is trying to figure out how these giant machines work, what their pieces look like, and how it all fits together. And so when the AlphaFold 2 model came out, he just had to try it.

• 8:33Speaker 07

  And I input a sequence, and then a few hours later, I had the model, and I was like, oh, my god. This just did it. And we'd been struggling with that problem for, for, you know, months, if not years.

• 8:48Speaker 07

  And all of a sudden, I messaged my lab and I said, we model everything.

• 8:54Speaker 07

  And we have, we've had dozens of projects that immediately progressed, uh, thanks to this.

• 8:59Emily Kwong?

  Okay. So it sounds like overnight, AlphaFold changed the trajectory of his lab.

• 9:04Speaker 02

  Yeah.

• 9:05Emily Kwong?

  But how did he know that using AlphaFold 2 would actually work?

• 9:09Speaker 02

  Yeah. So the accuracy is super important, right? Especially when you're basing all of your other experiments on the results. And it's important to note that like other AI, AlphaFold 2 isn't right 100% of the time. So you can't just take the results at face value. But unlike some other AI, included in the results is a score basically telling you how accurate each part of the structure is.

• 9:32Emily Kwong?

  Okay. And are others in the field using AlphaFold 2?

• 9:37Speaker 02

  Yeah. So this is something that actually sets AlphaFold apart from other protein prediction AI models. It's extremely user friendly. So essentially, anyone who works on a protein or even just has a sequence of a protein can plug it in and get results. I talked to Pushmeet Kohli, Vice President of Research at Google DeepMind, and he told me why it was important for them to make this tool open access.

• 10:00Speaker 08

  The mission statement that we have for the science program at Google DeepMind is to leverage AI to accelerate and advance science.

• 10:09Emily Kwong?

  Okay. So I'm scrolling through the AlphaFold website and I'm seeing scientists using this model for all kinds of things. They're working on malaria and cancer research, drug discovery, plastic eating enzymes.

• 10:20Speaker 02

  Mm hmm. And last week, DeepMind released a new version, AlphaFold 3, which can predict the 3D structure of proteins and other kinds of biomolecules that they attach to.

• 10:31Emily Kwong?

  Why are those other biomolecules important?

• 10:34Speaker 02

  Yes. So I know we talked about how much proteins are super important. I love them. But I have to admit, they rarely work alone.

• 10:44Speaker 02

  And if we actually want to know how biology works as a whole, we need to understand how proteins work with their partner molecules.

• 10:51Speaker 08

  So it really gives you a more detailed and more accurate picture of what is happening inside the body, where proteins are just, not just sort of existing in isolation, they are interacting in a very rich biological space or soup of RNA and DNA and small molecules. And it really sheds light into those rich interactions.

• 11:15Speaker 02

  Now, previous versions of these protein prediction softwares would model where each amino acid was located, but in this new version, AlphaFold 3, it maps things on an even smaller level. So it models where individual atoms are.

• 11:29Emily Kwong?

  Wow.

• 11:30Speaker 02

  So they can predict the structure of multi-protein complexes like the bacterial flagellum, or something like proteins in the blood which attach to iron atoms.

• 11:39Emily Kwong?

  That is powerful. Okay. What are the limits to AlphaFold predictions?

• 11:45Speaker 02

  Yeah, there are definitely limitations. Pushmeat says that the model works best when a protein has a single defined structure, but some proteins have more than one shape, or they have sections that are kind of flimsy. Think cooked versus uncooked spaghetti.

• 12:00Emily Kwong?

  Okay, so the model has, sounds like some trouble with prediction in some cases, and the results show that.

• 12:06Speaker 02

  Yeah, so the idea is that these results would say, hey, I'm not so confident in this area of the protein, just so like users know.

• 12:13Emily Kwong?

  Oh.

• 12:14Speaker 02

  And another limitation is that the prediction ability depends on the amount of what's called training data available.

• 12:21Emily Kwong?

  Uh huh.

• 12:22Speaker 02

  So I mentioned that there's a lot of training data for proteins, but...

• 12:26Speaker 08

  Some categories have much less training data available. For example, there's much less structural data available for RNAs.

• 12:36Emily Kwong?

  Okay, so the prediction is only as good as the data.

• 12:41Speaker 02

  Exactly, exactly. But Emily.

• 12:43Emily Kwong?

  But Burley.

• 12:43Speaker 02

  There's another way scientists can use AI in the protein world.

• 12:48Emily Kwong?

  Okay, what's that?

• 12:49Speaker 02

  To generate brand-new proteins, ones like not found in nature anywhere.

• 12:55Emily Kwong?

  Huh.

• 12:55Speaker 09

  Humans face new problems today, and you know, we live longer, we're polluting and heating up the planet, and it's reasonable to think that if with another million, more millions of years of evolution, that some of these problems would be, would be solved. But we don't want to wait that long. So the idea is that we can now create completely new proteins that solve these problems that weren't really relevant during evolution to make the world a better place.

• 13:22Speaker 02

  So this is David Baker. He's a biochemist and the director of the Institute for Protein Design at the University of Washington. And he's been working on proteins for years. He actually developed one of the earlier protein prediction models. His lab has a similar AI program to AlphaFold 3. It's called RosettaFold All-Atom. But his big focus is designing these brand-new proteins.

• 13:44Emily Kwong?

  This sounds so futuristic.

• 13:46Speaker 02

  Right?

• 13:47Emily Kwong?

  Like, what kind of new proteins?

• 13:50Speaker 02

  So far, they've done things like design new protein antibodies, which are important for fighting infections, in this case, to fight influenza. They've made something called a switch protein that could be used as an environmental sensor. And they've also made proteins that could help store carbon, which is a huge hurdle for fighting climate change.

• 14:09Speaker 09

  I think really across, you know, medicine, sustainability, technology, I think there's huge opportunities to transform the current ways we do things with protein design.

• 14:22Speaker 02

  So these predictive and generative AI models have fundamentally changed the protein science landscape. And again, there's definitely room for improving the prediction power. But with what the field has shifted to, like in terms of prediction accuracy and design potential, I mean, it's really gotten this retired protein fanatic, like, missing my science days.

• 14:48Emily Kwong?

  Burley, thank you so much for bringing us this big, big story about the little things in life.

• 14:56Speaker 02

  Thanks, Emily.

• 15:01Emily Kwong?

  This episode was produced by Rachel Carlson. It was edited by our showrunner, Rebecca Ramirez. Burley checked the facts. Co-Takasugi Chernovin was the audio engineer. Special thanks to Jeff Brumfield. Beth Donovan is our senior director. And Colin Campbell is our senior vice president of podcasting strategy. I'm Emily Kwong. Thank you for listening to Short Wave from NPR.

• 15:44Speaker 00

  Stay tuned for the following announcements and messages from our sponsors.