Artificial intelligence could help speed the drug-development process, researchers came up with an answer to a stem-cell riddle, and scientists in China designed a gel that could spur the regeneration of tooth enamel, giving hope to candy enthusiasts and coffee drinkers everywhere. Take a bite out of the future — confidently! — with this week’s coolest scientific discoveries.
What is it? Searching for ways to use artificial intelligence to speed up the drug-development process, the biotech startup Insilico Medicine announced that its AI program — called generative tensorial reinforcement learning, or GENTLR — had taken just 21 days to design six novel molecules that could be used to fight fibrosis and other diseases.
Why does it matter? As Wired reports, getting a single new drug to market can cost up to $2.6 billion, “and pipelines are only getting slower and more expensive.” A lot of footwork is required long before a drug even reaches the clinical-trial stage. The idea is that properly trained AI might shoulder some of that burden, acting as a kind of computer chemist by analyzing past successes and trying to formulate novel structures that could be created in the lab. Chemical biology professor Adam Renslo, who wasn’t involved in the research, said, “It’s cool to see AI trained to think a little bit like how a medicinal chemist thinks.”
How does it work? In 2016, Insilico published a paper discussing how a form of AI called generative adversarial networks, or GANs — commonly used to “generate images with specific properties” — could be used for drug discovery, and it’s been pursuing the idea ever since; the latest findings are in Nature Biotechnology. GENTLR, Wired explains, “works by looking at past research and patents for molecules that are known to be effective against … the particular drug target, as well as other structures. The idea is to prioritize novel, but logical structures, and those that could be synthesized in the lab.” Tasked with designing molecules to fight fibrosis, GENTLR designed six novel molecules in 21 days, one of which was tested in mice with promising results.
What is it? Researchers at Massachusetts Institute of Technology have built a microprocessor from carbon nanotube transistors — a big step toward creating the ultra-efficient computers of the future.
Why does it matter? Today’s silicon-based computer processors have had a good run, but as MIT News explains, they’re approaching a limit in terms of how small and efficient they can possibly be. Thus scientists have been looking at the next frontier, considered to be carbon nanotube field-effect transistors. CNFETs for short, they’ve got “properties that promise around 10 times the energy efficiency and far greater speeds compared to silicon.”
How does it work? The problem with carbon nanotube (CNT) microprocessors had been that, despite their promise, they were tricky to manufacture at scale: Essentially, “advanced circuits will need carbon nanotubes at around 99.999999% purity, which is virtually impossible to produce today.” The researchers worked around the problem by designing a technique called DREAM, for “designing resiliency against metallic CNTs,” which they explain in a new paper in Nature; with their method, the carbon nanotubes need only be at about 99.99% purity. Piece of cake!
What is it? A collaboration of researchers from the University of California, San Francisco; the National Heart, Lung, and Blood Institute; and Stanford University has shed light on why the immune system might reject stem cell treatments even if the cells come from the same body they’re being transplanted back into — and what can be done to make such treatments work better.
Why does it matter? As UCSF surgery professor Tobias Deuse, lead author of a new paper in Nature Biotechnology, explains in The Conversation, physicians and scientists have long sought to treat common diseases, including liver failure and Parkinson’s, by replacing failing cells with induced pluripotent stem cells (iPSCs), which can develop into any kind of cell in the body. The reasoning went that if those cells were sourced from the patient’s own body, the immune system wouldn’t reject them. That turned out not to be the case, though; as UCSF put it in a press release, “iPSCs haven’t emerged as the cure-all that was originally envisioned,” and scientists have been trying to understand why.
How does it work? The new study finds that when adult cells are converted to iPSCs, the process can “mutate DNA found in tiny cellular structures called mitochondria,” which can then trigger an immune response when the cells are transplanted back to the body. According to Deuse, that gives researchers a lead for figuring out a way around the problem: “To dodge the immune system and make regenerative stem cell therapies widely available to the general public, our lab aims to engineer stem cells lacking any immune features.” They can do so, he adds, with modern gene-editing technology, which has already shown success in the lab with both edited mouse and human stem cells.
What is it? Scientists at Rice University designed a catalytic reactor that turns carbon dioxide into formic acid, a kind of liquid fuel.
Why does it matter? The process was developed in the lab of chemical and biomolecular engineer Haotian Wang, who said that the formic acid could have many uses, besides being a way to repurpose a potent greenhouse gas. For one, it’s an “energy carrier,” Wang said. “It’s a fuel-cell fuel that can generate electricity and emit carbon dioxide — which you can grab and recycle again. It’s also fundamental in the chemical engineering industry as a feedstock for other chemicals, and a storage material for hydrogen that can hold nearly 1,000 times the energy of the same volume of hydrogen gas, which is difficult to compress. That’s currently a big challenge for hydrogen fuel-cell cars.”
How does it work? The device relies on two new advances — one is a solid-state electrolyte that eliminates the need for salt in the process, thereby making this method cheaper and less energy-intensive than other technologies to reduce carbon dioxide; the other is the use of bismuth, a heavy atom that stabilizes the catalyst. The findings are explained further in a paper in Nature Energy. “The big picture is that carbon dioxide reduction is very important for its effect on global warming as well as for green chemical synthesis,” Wang said. “If the electricity comes from renewable sources like the sun or wind, we can create a loop that turns carbon dioxide into something important without emitting more of it.”
What is it? In China, researchers at Zhejiang University developed a gel that spurs the regeneration of tooth enamel — the hardest biological tissue, and one that hasn’t been duplicated artificially, according to a new paper in Science Advances.
Why does it matter? Tooth enamel has a complicated structure and can’t repair itself — which is especially unfortunate given how much abuse humans visit upon our teeth, what with all the soda, gummy bears and so forth that people consume. Degradation of this protective enamel is what leads to cavities.
How does it work? Trigger warning for the squeamish: “The scientists tested the product on damaged human teeth that had been removed from patients and kept in a solution that recreates the mouth environment,” according to the Independent. The substance, which the scientists achieved by “mixing calcium and phosphate ions — two minerals which are found in the enamel — with the chemical trimethymaline in an alcohol solution,” caused enamel to grow on the teeth with the same structure as natural enamel, adding 2.5 micrometers of thickness within 48 hours. As a release from the university notes, the substance has not yet been tried in the “hostile environment” of an actual human mouth. Professor Tang Ruikang, who led the research team, reportedly has cracked enamel himself; he might be the first test subject.