School of Science

At a recent conference, Prof. Sergey Ovchinnikov and his colleagues presented a paper demonstrating how they have used advanced versions of ChatGPT to “generate amino acid sequences that code for biologically active proteins with a structural feature called a four-helix bundle,” reports Jeffrey Brainard for Science. “To Ovchinnikov’s surprise, ChatGPT produced gene sequences without further refinement of his team’s query,” writes Brainard. “Still, the application of ChatGPT to this task needs refinement, Ovchinnikov found. Most of the sequences his team produced did not garner 'high confidence' on a score predicting whether they would form the desired protein structure.” 

Prof. Alex Shalek and his colleagues developed a deep-learning model called DrugReflector aimed at speeding up the process of drug discovery, reports Heidi Ledford for Nature. “They used DrugReflector to find chemicals that can affect the generation of platelets and red blood cells — a characteristic that could be useful in treating some blood conditions,” explains Ledford. The researchers found that “DrugReflector was up to 17 times more effective at finding relevant compounds than standard, brute-force drug screening that depends on randomly selecting compounds from a chemical library.”

George Smoot '66, PhD '70, recipient of the 2006 Nobel Prize in Physics, has died at the age of 80, reports Katrina Miller for The New York Times. Smoot’s work as a physicist “helped elucidate the story of cosmic creation, providing evidence of what he called the primordial seeds that grew into galaxies and galaxy clusters,” writes Miller. 

Researchers at MIT have uncovered remnants of “proto Earth,” which existed before a Mars-sized meteorite slammed into the Earth billions of years ago, reports Gayoung Lee for Gizmodo. The researchers “found an odd imbalance of potassium isotopes in ancient rock samples,” explains Lee. “Chemical analyses revealed the anomaly couldn’t have emerged from any known geological processes on modern Earth.” 

New York Times reporter Annabel Keenan highlights the “Remembering the Future” exhibit at the MIT Museum, a sculptural installation created by Janet Echelman that uses “climate data from the last ice age to the present, as well as projected future environments, to create a geometric design.” Echelman worked with MIT faculty, including Prof. Raffaele Ferrari and Prof. Caitlin Mueller, to bring the project to life. Mueller explains that she developed a “high-fidelity digital twin of the sculpture generated through our computational simulation that you can orbit and pan through to get perspectives that you can’t see physically in the space.” 

Prof. Peter Shor speaks with Katia Moskvitch and her nine-year-old son Kai of The Quantum Kid about how quantum technologies could be applied to developing climate change and sustainability solutions. Shor explains that quantum computers can be used for “simulating quantum mechanics, which would really help immensely in designing new materials, and new materials could be very useful for solving the climate crisis.” 

Graduate student Chung-Tao (Josh) Chou speaks with Tech Briefs reporter Andrew Corselli about his work developing a magnetic transistor that could lead to more energy-efficient circuits. “People have known about magnets for thousands of years, but there are very limited ways to incorporate magnetism into electronics,” says Chou. “We have shown a new way to efficiently utilize magnetism that opens up a lot of possibilities for future applications and research.”

Using molecular evidence buried in rocks, researchers at MIT suggest that some of the Earth’s first living creatures are ancestors of the modern sea sponge, reports Ashley Fike for Vice. “The discovery suggests the earliest animals were simple, filter-feeding organisms that slowly cleaned the seas while the rest of the evolution was still figuring itself out,” says Fike. “These early sponges likely had no skeletons, nerves, or eyes – just porous bodies that absorbed water and nutrients. Yet they paved the way for everything that came next, from insects to mammals to us.” 

Writing for Forbes, Senior Lecturer Guadalupe Hayes-Mota '08, SM '16, MBA '16 emphasizes the importance of implementing ethical frameworks when developing AI systems designed for use in healthcare. “The future of AI in healthcare not only needs to be intelligent,” writes Hayes-Mota. “It needs to be trusted. And in healthcare, trust is the ultimate competitive edge.” 

George Smoot '66, PhD '70, the 2006 recipient of the Nobel Prize in Physics, has died at the age of 80, reports Kasha Patel for The Washington Post. Smoot’s work “helped take an image of the universe in its infancy, providing strong support for the Big Bang theory and new insight into the origins of the cosmos,” writes Patel. 

MIT researchers have developed a new model that illustrates the chemical mechanisms underlying lithium-ion batteries, reports Gayoung Lee for Gizmodo. The findings could “lead to faster, more efficient batteries for electric vehicles, portable electronics, and more.”

A new study by MIT researchers suggests sea sponges may have been the “first animals to inhabit the Earth,” reports Maria Azzura Volpe for Newsweek. “In their work, the researchers linked so-called ‘chemical fossils’ found in ancient rocks to the ancestors of a class of modern-day sea sponges known as demosponges,” explains Volpe. “These chemical fossils—the molecular remnants of once-living organisms that have been buried, transformed, and preserved in sediment over time—were discovered in rocks that date back to more than 541 million years ago, during the Ediacaran Period.” 

MIT researchers have uncovered new evidence that suggests some of Earth’s first living creatures are ancestors of the modern sea sponge, reports Andrew Paul for Popular Science. The researchers identified 541 million-year-old chemical fossils embedded in sediment that they believe may indicate that some of Earth’s earliest creatures were the ancient relatives of today’s sea sponges. 

George Smoot '66, PhD '70, a recipient of the 2006 Nobel Prize in Physics for his work “finding the background radiation that finally pinned down the Big Bang Theory,” has died at age 80, reports the Associated Press. 

In an effort to better understand how protein language models (PLMs) think and better judge their reliability, MIT researchers applied a tool called sparse autoencoders, which can be used to make large language models more interpretable. The findings “may help scientists better understand how PLMs come to certain conclusions and increase researchers’ trust in them," writes Andrea Luis for The Scientist.