Artificial intelligence could help seismologists predict earthquakes, a new type of drug could flush aging cells from the body, and researchers introduced a robot that’s made of … smaller robots. Similarly, this week’s roundup of cool scientific discoveries is way more than the sum of its parts.
What is it? A team led by Paul Johnson, a geophysicist at Los Alamos National Laboratory, is investigating ways to predict earthquakes using artificial intelligence — specifically, machine learning algorithms that can identify patterns in troves of data. Their new paper, published on arXiv.org, hasn’t been peer-reviewed yet but shows “tantalizing” results, according to an article in Quanta magazine.
Why does it matter? Predicting where and when an earthquake might hit would be helpful for obvious reasons, though some seismologists don’t think it can be done. The stakes are enormous, particularly in places due for a really big one, like the Cascadia subduction zone — where a tectonic plate beneath the Pacific Ocean is slowly wedging itself underneath North America. “In the best-case scenario,” writes Quanta’s Ashley Smart, “predictions of big earthquakes will probably have time bounds of weeks, months or years. Such forecasts probably couldn’t be used, say, to coordinate a mass evacuation on the eve of a temblor. But they could increase public preparedness, help public officials target their efforts to retrofit unsafe buildings, and otherwise mitigate hazards of catastrophic earthquakes.”
How does it work? Johnson et al used blocks to create “laboratory earthquakes,” then fed various data points, including acoustic signals, into an algorithm, which analyzed the data to identify signs a quake might be coming. They then applied what they’d learned to real-world “slow slip” earthquakes — more routine seismic movements beneath the earth’s surface that we typically can’t feel, but that may increase stress at the earth’s crust, where disastrous earthquakes take place. Though it didn’t do quite as well outside the lab as in it, the AI was able to predict most slow slip quakes it trained its analysis on to within a matter of days. Maarten de Hoop, a Rice University seismologist who wasn’t involved with the research, told Quanta that the results were a start: “Machine learning techniques have given us a corridor, an entry into searching in data to look for things that we have never identified or seen before.”
What is it? Bacteria that are resistant to antibiotics are a major public health challenge — but what if the genes that make those bacteria drug-resistant could simply be removed? That’s the idea behind new research published in Antimicrobial Agents and Chemotherapy, which could give immunologists new ammo for fighting out-of-control infections.
Why does it matter? The World Health Organization calls antibiotic resistance “one of the biggest threats to global health, food security, and development today.” Due to both human and animal overuse of antibiotics, which has led them to become less effective on many bacteria, it’s becoming harder to treat infections including tuberculosis, pneumonia and gonorrhea. WHO concludes, “Without urgent action, we are heading for a post-antibiotic era, in which common infections and minor injuries can once again kill.”
How does it work? A team of scientists in Colorado and Texas used the genome-editing system CRISPR-Cas9 to engineer a plasmid — a small DNA fragment inside a cell but outside its chromosomes. Plasmids can replicate on their own and provide bacteria with useful tools, like antibiotic resistance. The team engineered their plasmid to remove the gene that enables resistance to antibiotics in Enterococcus faecalis, a normal part of the intestinal flora that can become pathogenic if antibiotics throw things out of whack. In hospital settings, infection related to E. faecalis is a “major problem,” according to the American Society for Microbiology. “The engineered plasmid can significantly reduce the occurrence of antibiotic resistance in the target bacterial population, rendering it more susceptible to antibiotics,” said co-senior author Breck A. Duerkop, of the University of Colorado School of Medicine. In mouse models, it reduced the resistance gene threefold.
What is it? In a small clinical trial, researchers at Minnesota’s Mayo Clinic used a type of drug called senolytics to clear old and decaying cells from the bodies of human patients — an advance in a promising field of treatments to combat the effects of aging.
Why does it matter? Senolytics, an emerging and “highly anticipated” field, focuses on senescent cells — “malfunctioning cells that accumulate with aging and in organs affected by chronic diseases,” according to the Mayo Clinic. “Senescent cells can remain in the body and contribute to multiple diseases as well as features of aging, ranging from heart disease to frailty, dementias, osteoporosis, diabetes, and kidney, liver, and lung diseases.” James Kirkland, the senior author of a new paper published in EBioMedicine, said that earlier trials in mice showed that senolytics could “delay, prevent, or treat multiple diseases and increase health and independence during remaining years of life.”
How does it work? With the earlier mouse trials, the proof was in the pictures, according to the Guardian: Mayo released photos of two elderly mice, one of which had been treated with senolytics and looked a heck of a lot younger than its wizened counterpart. The new clinical study, which involved patients with diabetes-related kidney disease, was aimed to demonstrating that senolytics can safely be used in humans to clear senescent cells from the body; more work is needed, but it’s “a significant step forward for translation of senolytic therapies,” said the National Institute of Aging’s Ronald Kohanski.
What is it? Researchers at Georgia Tech and Northwestern University created a robot made of smaller 3D-printed robots — locomotive components called smarticles, for “smart active particles.”
Why does it matter? Typical robots require many components designed to fit just so, with complex motor and battery systems to make all the parts work together; the researchers’ vision of a simpler system could signal a new paradigm in robotics, particularly small robotics. “These are very rudimentary robots whose behavior is dominated by mechanics and the laws of physics,” said Georgia Tech physics professor Dan Goldman. “We are not looking to put sophisticated control, sensing, and computation on them all. As robots become smaller and smaller, we’ll have to use mechanics and physics principles to control them because they won’t have the level of computation and sensing we would need for conventional control.”
How does it work? The smarticles do only one thing: They flap their tiny arms back and forth. But more than one smarticle together, confined in a circle, can form a system that moves by itself, called a “supersmarticle,” and that could be directed by light or sound stimulus enough to successfully navigate a maze. The researchers described their smarticle accelerator in a paper in Science Robotics.
What is it? Researchers from Aalto University and VTT Technical Research Centre of Finland took cues from wood and spider silk to create a new bio-based material that could “rival plastic,” combining strength and extensibility — that is, stretchiness.
Why does it matter? The material has many possible uses as a replacement for plastic, according to a release from Aalto University, including “in medical applications, surgical fibres, the textile industry and packaging.” But it’s also biodegradable — unlike plastic — and wouldn’t lead to the mounting ecological damage caused by microplastics.
How does it work? Taking their cue from nature, the researchers combined wood cellulose fibers with a silk protein found in spider webs, explained VTT’s Pezhman Mohammadi: “We used birch tree pulp, broke it down to cellulose nanofibrils and aligned them into a stiff scaffold. At the same time, we infiltrated the cellulosic network with a soft and energy dissipating spider silk adhesive matrix.” The silk protein was produced in the lab from synthetic DNA. The bio-material is described further in Science Advances.