56 Artificial Intelligence Examples to Know for 2024
AlphaFold 3 Unlocks A New Scientific Era, Mastering 'all Of Life's Molecules'
Proteins are one of nature's most incredible and versatile inventions. These essential building blocks of life catalyze virtually every chemical reaction in the body, regulate gene expression and the immune system, make up the major structural elements of all cells, and form the major components of muscle.
But why's this AI called 'AlphaFold'? Proteins are chains of amino acids that fold spontaneously to form a 3D structure crucial to the protein's biological function. You can look at their components and sequences on paper, but if you don't know what 3D shape they fold up into, you can't predict what they'll do, or how they'll interact with other molecules.
Various human-operated experimental techniques have been used to determine protein structures, but they're time-consuming and expensive. Considering that there are over 200 million known proteins across all lifeforms – and, so far, only 170,000 protein structures have been identified – it made sense to co-opt AI to speed up the process. But, until AlphaFold, programs haven't been able to predict protein structures in a way that's as accurate as human-based experimental techniques.
AlphaFold 2, released in 2021, was a game-changing breakthrough, predicting 3-D structures for nearly every protein in the human body and enabling all sorts of cutting-edge science. In less than three years, it's been used by researchers worldwide to accelerate discoveries in cancer treatment, malaria vaccines, the creation of plastic-eating enzymes, and countless others – the Alphafold 2 paper currently lists more than 14,000 citations.
So, how is AlphaFold 3 an improvement? Well, the newest version moves beyond just predicting the structure and interactions of proteins to include everything from multiple proteins to DNA, RNA, and small-molecule ligands. (Most drugs are ligands that bind to proteins to change how they interact in human health and disease.)
This makes it an absolutely unprecedented resource for simulating how specific proteins in the body will interact with specific drug molecules.
3D protein structure of the spike protein of a common cold virus generated by AlphaFold 3Google DeepMind/Isomorphic Labs
The GIF above gives you an idea of the structure that AlphaFold 3 generates. It shows its structural prediction for a spike protein (blue) of a common cold virus as it interacts with antibodies (turquoise) and simple sugars (yellow). You can see how accurately AlphaFold 3's model matches the protein's true structure, which is underlaid in gray.
To achieve these advanced capabilities, AlphaFold 3 was trained on global molecular structural data held within the Protein Data Bank. Deepmind says it can process over 99% of all known biomolecular complexes in that database. In addition, its Evoformer module, the architecture that allowed AlphaFold 2 to perform like it did, was improved.
This is how the AlphaFold system works in very, very basic terms. (Thanks to the University of Oxford's Protein Informatics Group for its easier-to-understand explanation.) It takes the amino acid sequence that's input, searches databases for similar sequences already identified in other living organisms, extracts all of the relevant information using a transformer (that's the Evoformer), and passes that information on to a neural network that produces a 3D structure – a long list of coordinates representing the position of each atom of the protein, including side chains.
AlphaFold 3's prediction for a molecular complex featuring a protein (blue) bound to a DNA double helix (pink) is a near-perfect match to the true molecular structure (gray)Google DeepMind/Isomorphic Labs
What the new-and-improved Evoformer does is assemble its structural predictions using a diffusion network, like those found in AI image generators. As the joint blog post by DeepMind and Isomorphic Labs announcing AlphaFold 3 explains, it "starts with a cloud of atoms, and over many steps converges on its final, most accurate molecular structure."
In a recent interview with Bloomberg's Tom Mackenzie, Google DeepMind CEO and co-founder (and CEO and founder of Isomorphic Labs) Demis Hassabis discussed the implications of using AlphaFold 3 in drug discovery.
"The holy grail of drug discovery is not just knowing the protein structure, which is what AlphaFold 2 did, but actually designing drug compounds called ligands that bind to the protein's surface," Hassabis said. "And you want to know where it binds, and how strongly it binds, in order for you to design the right kind of drug compound. So, AlphaFold 3 is a big step in that direction of predicting protein-ligand binding and how that interaction will work."
In January this year, Isomorphic Labs announced a strategic partnership with pharmaceutical giants Eli Lilly and Novartis, worth a combined value of around US$3 billion. But what's incredible here is the drug production timeline that's expected to result from these partnerships.
"So, we're already working on real drug programs," said Hassabis. "And I would be expecting maybe in the next couple of years the first AI-designed drugs [appearing] in the clinic."
"If you ask me the number one thing AI could do for humanity," he continued, "it would be to solve, you know, hundreds of terrible diseases. I can't imagine a better use case for AI. So that's partly the motivation behind Isomorphic and AlphaFold and all the work we do in science, it's to advance society in these big ways."
The full interview between Hassabis and Mackenzie can be viewed in the video below.
Google DeepMind CEO on Drug Discovery, Hype, Isomorphic
When it was tested, AlphaFold 3 demonstrated state-of-the-art accuracy in predicting drug-like interactions, including proteins bound with ligands and antibodies bound with target proteins.
Using the PoseBusters benchmark, it was found to be 50% more accurate than the best existing methods – without the need to input any structural information. PoseBusters checks the chemical and physical plausibility of molecular and protein-ligand 'poses' generated by a deep-learning model.
And you can play with it yourself. AlphaFold 3 is available via the AlphaFold Server, which includes a database of 200 million protein structures. This phenomenal resource is free to scientists conducting non-commercial research – or indeed just curious Web users worldwide.
AlphaFold Server Demo - Google DeepMind
Predicting protein structures without a tool like this can take... Well, about as long as it takes to complete a PhD, and can cost hundreds of thousands of dollars. Much like how DeepMind's GNoME tool has catapulted materials and crystals discovery hundreds of years into the future, AlphaFold 3 promises to radically accelerate vast areas in biological science and pharma.
"This new window on the molecules of life reveals how they're all connected and helps understand how those connections affect biological functions – such as the actions of drugs, the production of hormones and the health-preserving process of DNA repair," said Google DeepMind and Isomorphic Labs. "This leap could unlock more transformative science, from developing biorenewable materials and more resilient crops, to accelerating drug design and genomics research."
We can't wait to see where this technology takes us.
Research into the predictive capabilities of AlphaFold 3 was published in Nature.
Sources: Google DeepMind, Isomorphic Labs
Innovations In Recycling
Plastic waste is one of the most urgent environmental issues of our time. Less than 10 percent of the plastic we use is recycled, and there's currently an estimated 100 million tons of plastic in oceans around the world. But what would happen if we stopped thinking of plastic as waste, and instead as a valuable renewable resource?
Scientists around the world want to find out. The plastic "end-of-life challenge" calls for new ways to recycle and reuse plastics endlessly in a closed loop system, so they never become waste. Innovation on that scale would convert the current "make-take-dispose" linear economy into a circular economy, where recycling plastic for eternity is possible.
One scientist has made a significant advancement. John Layman, head of material science at Procter & Gamble and chief technologist and founding inventor of PureCycle Technologies, developed a revolutionary process to remove color, odor, and contaminants from polypropylene plastic waste and transform it into a "virgin-like" resin, which is the basis for plastic products. PureCycle's technology presents a major development in recycling capabilities, and focusing on polypropylene is especially notable. It's the second-most used plastic in the world, yet only 1 percent is currently recycled.
PureCycle plans to recycle, purify, and modify difficult materials like this carpet waste recovered by Circular Polymers in California.
Photograph Courtesy Care Recycling (Top) (Left) and Photograph Courtesy Care Recycling (Bottom) (Right)
Before and after: Black polypropylene pellets made from recycled carpet fibers are purified and transformed into clear plastic resin through the PureCycle process.
Photograph Courtesy Milliken
Layman's colleague and former classmate Scott Trenor, a senior polymer scientist at Milliken & Company, contributed a key set of plastic additives to increase the viability of PureCycle materials. Additives are chemical substances that modify the properties of plastics so they can be used in different types of products. For example, a car bumper would need to be more durable and impact-resistant, while a yogurt cup would need to be more flexible. Now Milliken and PureCycle are working together to scale and advance the technology, with plans to start commercial-scale production at PureCycle's first plant, in Ohio, in 2021.
Trenor says the collaboration is a natural fit, partly because "both companies have a very strong environmental purpose." Milliken's first recycling policy dates back to 1901, and the company just launched ambitious sustainability goals that include creating circular economies and zero waste to landfill by 2025. PureCycle's technology supports P&G's vision of using 100 percent recyclable or renewable materials in its packaging.
Trenor's motivation is closer to home. He often takes his 4-year-old son to ride mountain bikes near his home in South Carolina, and he's been disheartened to see an increase in plastic litter in natural spaces. "It's not the way I want to leave things for the future generation," he says.
Layman agrees. While studying plastics in graduate school, he was amazed at the volume of material being produced, and horrified by how much was ending up in the environment. "Since the day you were born, there's a pile of trash with your name on it," he says. "The first diaper you wore as a newborn is likely still on this planet somewhere."
PureCycle aims to close the loop on the plastics lifecycle, converting the current "make-take-dispose" linear economy into a circular economy, where plastic waste never leaks into the environment.
Photograph Courtesy Milliken
He got interested in recycling in 2008, when he was tasked with buying plastic waste from recyclers and surveying its usefulness for P&G's products and packaging. "You quickly realize there's a lot of issues when it comes to the quality of recycled material," he says. Currently, only two kinds of plastic, PET and HDPE, are economically viable for recyclers, and even those are hard to upcycle into high-quality products.
Layman focused on polypropylene because it's one of the three largest plastic resins used in the world. Its super powers include flexibility and impact resistance. It's found in most caps on most bottles. It's in luggage and carpets, computers and phones. In the grocery store, it's everywhere—yet it's hardly the favorite of recyclers looking to make a profit. It holds onto pungent smells and contaminants, and it can only be made into black or gray products. For those reasons, the little that's recycled is usually made into park benches or car bumpers—important but limited applications.
To recycle polypropylene into higher-value products, Layman knew he would first have to purify the plastic waste, and in an energy-efficient way. He worked on the discovery phase with financial backing from an internal seed fund program at P&G. The resulting PureCycle technology relies on a physical solvent-based process that uses less energy than a chemical process because it doesn't have to break down and build up the molecule. "It's the combination of the solvent choice, plus specific process steps, that enable us to purify this material in a way that nobody's been able to do before," he says.
PureCycle's first plant in Ironton, Ohio, is expected to purify and recycle 119 million pounds of polypropylene and produce 105 million pounds each year. Commercial production will begin in 2021.
Photograph Courtesy Milliken (Top) (Left) and Photograph Courtesy Milliken (Bottom) (Right)
Once the material has been purified, the question is what to do with it. Trenor explains that when scientists create plastic products from scratch, "they have a lot of knobs they can turn to select the exact set of properties they need for a certain application." Working with recycled material is more challenging.
That's where Milliken's additives come in—they can modify PureCycle's polypropylene resin for use in a diverse set of applications. For the first time, recycled polypropylene doesn't have to become a car bumper. Purified and modified, the resin can be molded into different products with different properties in a closed loop.
ENVIRONMENT
Your hair is surprisingly recyclableTRAVEL
Kit list: the essential recycled travel gear, from beachwear to backpacksENVIRONMENT
Why your recycling doesn't always get recycledThe first PureCycle plant is expected to purify and recycle 119 million pounds of polypropylene and produce 105 million pounds each year. Those numbers sound huge, but Layman puts it in perspective by pointing out that 120 billion pounds of polypropylene were produced globally in 2018 alone. "You can see we have a long way to go," he says.
He compares PureCycle with wind and solar energy technology before they scaled up. "We have an ambition to build 25 plants around the world, each one bigger than the last," he says. "This is plant number one." Eventually, he hopes to PureCycle at least 10 to 20 percent of all polypropylene plastic.
"For all of these technologies, it's really more of a marathon than a sprint," says Trenor. He also notes that the plastic end-of-life challenge doesn't concern technology alone. For PureCycle or any other initiative to succeed, consumers need to change their behavior and recycle more, and recyclers need the ability and financial incentive to process more than PET and HDPE.
Recycled post-consumer waste is ready for sorting.
Video Still Courtesy Pacific Northwest Secondary Sorting Demonstration Project
Recyclers use tools like Near-Infrared Sorting to separate different types of materials in the recycling stream.
Video Still Courtesy Pacific Northwest Secondary Sorting Demonstration Project
A bale of sorted polypropylene waste is ready for buyers. The Association of Plastic Recyclers reported in 2012 that in North America alone, there was demand for 1 billion pounds of recycled polypropylene.
Photograph Courtesy Association of Plastic Recyclers
Trenor says that U.S. Recycling facilities are part of the problem. "A lot of the separation capabilities in their infrastructure was built 20 or 30 years ago when there wasn't as much plastic," he explains. "So they were built for paper, cardboard, steel, and glass. As those materials have decreased in volume and plastics have increased, the technology and infrastructure hasn't kept up."
Meanwhile, demand for recycled plastic is high. Consumer goods manufacturers are committing to using minimum percentages of recycled content in their packaging, and new legislation in California and Europe requires it. The Association of Plastic Recyclers reported in 2012 that in North America alone, there was demand for 1 billion pounds of recycled polypropylene.
Supply, however, can't keep up. "Recyclers are screaming for more material because they can't fill their current orders," Trenor says.That's mainly because people just aren't recycling enough. "Once it's in the consumer's hand, there needs to be some personal responsibility to put it in the right bin, sort it in the right way, and make sure it doesn't end up in a landfill, or even worse, on a beach or in the forest somewhere."
Trenor is excited about his role in new innovations that have the potential to curb plastic waste, but he knows it will take more than science. "Milliken isn't going to solve this problem alone. PureCycle isn't going to solve this problem alone," he says. "We need a diverse group of people, companies, NGOs, and governments to work together to solve the problem of plastic waste."
Cellular Activity Hints That Recycling Is In Our DNA
Although you may not appreciate them, or have even heard of them, throughout your body, countless microscopic machines called spliceosomes are hard at work. As you sit and read, they are faithfully and rapidly putting back together the broken information in your genes by removing sequences called "introns" so that your messenger RNAs can make the correct proteins needed by your cells.
Introns are perhaps one of our genome's biggest mysteries. They are DNA sequences that interrupt the sensible protein-coding information in your genes, and need to be "spliced out." The human genome has hundreds of thousands of introns, about 7 or 8 per gene, and each is removed by a specialized RNA protein complex called the "spliceosome" that cuts out all the introns and splices together the remaining coding sequences, called exons. How this system of broken genes and the spliceosome evolved in our genomes is not known.
Over his long career, Manny Ares, UC Santa Cruz distinguished professor of molecular, cellular, and developmental biology, has made it his mission to learn as much about RNA splicing as he can.
"I'm all about the spliceosome," Ares said. "I just want to know everything the spliceosome does -- even if I don't know why it is doing it."
In a new paper published in the journal Genes and Development, Ares reports on a surprising discovery about the spliceosome that could tell us more about the evolution of different species and the way cells have adapted to the strange problem of introns. The authors show that after the spliceosome is finished splicing the mRNA, it remains active and can engage in further reactions with the removed introns.
This discovery provides the strongest indication we have so far that spliceosomes could be able to reinsert an intron back into the genome in another location. This is an ability that spliceosomes were not previously believed to possess, but which is a common characteristic of "Group II introns," distant cousins of the spliceosome that exist primarily in bacteria.
The spliceosome and Group II introns are believed to share a common ancestor that was responsible for spreading introns throughout the genome, but while Group II introns can splice themselves out of RNA and then directly back into DNA, the "spliceosomal introns" that are found in most higher-level organisms require the spliceosome for splicing and were not believed to be reinserted back into DNA. However, Ares's lab's finding indicates that the spliceosome might still be reinserting introns into the genome today. This is an intriguing possibility to consider because introns that are reintroduced into DNA add complexity to the genome, and understanding more about where these introns come from could help us to better understand how organisms continue to evolve.
Building on an interesting discovery
An organism's genes are made of DNA, in which four bases, adenine (A), cytosine (C), guanine (G) and thymine (T) are ordered in sequences that code for biological instructions, like how to make specific proteins the body needs. Before these instructions can be read, the DNA gets copied into RNA by a process known as transcription, and then the introns in that RNA have to be removed before a ribosome can translate it into actual proteins.
The spliceosome removes introns using a two-step process that results in the intron RNA having one of its ends joined to its middle, forming a circle with a tail that looks like a cowboy's "lariat," or lasso. This appearance has led to them being named "lariat introns." Recently, researchers at Brown University who were studying the locations of the joining sites in these lariats made an odd observation -- some introns were actually circular instead of lariat shaped.
This observation immediately got Ares's attention. Something seemed to be interacting with the lariat introns after they were removed from the RNA sequence to change their shape, and the spliceosome was his main suspect.
"I thought that was interesting because of this old, old idea about where introns came from," Ares said. "There is a lot of evidence that the RNA parts of the spliceosome, the snRNAs, are closely related to Group II introns."
Because the chemical mechanism for splicing is very similar between the spliceosomes and their distant cousins, the Group II introns, many researchers have theorized that when the process of self-splicing became too inefficient for Group II introns to reliably complete on their own, parts of these introns evolved to become the spliceosome. While Group II introns were able to insert themselves directly back into DNA, however, spliceosomal introns that required the help of spliceosomes were not thought to be inserted back into DNA.
"One of the questions that was sort of missing from this story in my mind was, is it possible that the modern spliceosome is still able to take a lariat intron and insert it somewhere in the genome?" Ares said. "Is it still capable of doing what the ancestor complex did?"
To begin to answer this question, Ares decided to investigate whether it was indeed the spliceosome that was making changes to the lariat introns to remove their tails. His lab slowed the splicing process in yeast cells, and discovered that after the spliceosome released the mRNA that it had finished splicing introns from, it hung onto intron lariats and reshaped them into true circles. The Ares lab was able to reanalyze published RNA sequencing data from human cells and found that human spliceosomes also had this ability.
"We are excited about this because while we don't know what this circular RNA might do, the fact that the spliceosome is still active suggests it may be able to catalyze the insertion of the lariat intron back into the genome," Ares said.
If the spliceosome is able to reinsert the intron into DNA, this would also add significant weight to the theory that spliceosomes and Group II introns shared a common ancestor long ago.
Testing a theory
Now that Ares and his lab have shown that the spliceosome has the catalytic ability to hypothetically place introns back into DNA like their ancestors did, the next step is for the researchers to create an artificial situation in which they "feed" a DNA strand to a spliceosome that is still attached to a lariat intron and see if they can actually get it to insert the intron somewhere, which would present "proof of concept" for this theory.
If the spliceosome is able to reinsert introns into the genome, it is likely to be a very infrequent event in humans, because the human spliceosomes are in incredibly high demand and therefore do not have much time to spend with removed introns. In other organisms where the spliceosome isn't as busy, however, the reinsertion of introns may be more frequent. Ares is working closely with UCSC Biomolecular Engineering Professor Russ Corbett-Detig, who has recently led a systematic and exhaustive hunt for new introns in the available genomes of all intron-containing species that was published in the journal Proceedings of the National Academy of Sciences (PNAS) last year.
The paper in PNAS showed that intron "burst" events far back in evolutionary history likely introduced thousands of introns into a genome all at once. Ares and Corbett-Detig are now working to recreate a burst event artificially, which would give them insight into how genomes reacted when this happened.
Ares said that his cross-disciplinary partnership with Corbett-Detig has opened the doors for them to really dig into some of the biggest mysteries about introns that would probably be impossible for them to understand fully without their combined expertise.
"It is the best way to do things," Ares said. "When you find someone who has the same kind of questions in mind but a different set of methods, perspectives, biases, and weird ideas, that gets more exciting. That makes you feel like you can break out and solve a problem like this, which is very complex."
Comments
Post a Comment