Transistors, the building blocks of modern electronics, are typically made of silicon. Because it’s a semiconductor, this material can control the flow of electricity in a circuit. But silicon has fundamental physical limits that restrict how compact and energy-efficient a transistor can be.

MIT researchers have now replaced silicon with a magnetic semiconductor, creating a magnetic transistor that could enable smaller, faster, and more energy-efficient circuits. The material’s magnetism strongly influences its electronic behavior, leading to more efficient control of the flow of electricity. 

The team used a novel magnetic material and an optimization process that reduces the material’s defects, which boosts the transistor’s performance.

The material’s unique magnetic properties also allow for transistors with built-in memory, which would simplify circuit design and unlock new applications for high-performance electronics.

“People have known about magnets for thousands of years, but there are very limited ways to incorporate magnetism into electronics. We have shown a new way to efficiently utilize magnetism that opens up a lot of possibilities for future applications and research,” says Chung-Tao Chou, an MIT graduate student in the departments of Electrical Engineering and Computer Science (EECS) and Physics, and co-lead author of a paper on this advance.

Chou is joined on the paper by co-lead author Eugene Park, a graduate student in the Department of Materials Science and Engineering (DMSE); Julian Klein, a DMSE research scientist; Josep Ingla-Aynes, a postdoc in the MIT Plasma Science and Fusion Center; Jagadeesh S. Moodera, a senior research scientist in the Department of Physics; and senior authors Frances Ross, TDK Professor in DMSE; and Luqiao Liu, an associate professor in EECS, and a member of the Research Laboratory of Electronics; as well as others at the University of Chemistry and Technology in Prague. The paper appears today in Physical Review Letters.

Overcoming the limits

In an electronic device, silicon semiconductor transistors act like tiny light switches that turn a circuit on and off, or amplify weak signals in a communication system. They do this using a small input voltage.

But a fundamental physical limit of silicon semiconductors prevents a transistor from operating below a certain voltage, which hinders its energy efficiency.

To make more efficient electronics, researchers have spent decades working toward magnetic transistors that utilize electron spin to control the flow of electricity. Electron spin is a fundamental property that enables electrons to behave like tiny magnets.

So far, scientists have mostly been limited to using certain magnetic materials. These lack the favorable electronic properties of semiconductors, constraining device performance.

“In this work, we combine magnetism and semiconductor physics to realize useful spintronic devices,” Liu says.

The researchers replace the silicon in the surface layer of a transistor with chromium sulfur bromide, a two-dimensional material that acts as a magnetic semiconductor.

Due to the material’s structure, researchers can switch between two magnetic states very cleanly. This makes it ideal for use in a transistor that smoothly switches between “on” and “off.”

“One of the biggest challenges we faced was finding the right material. We tried many other materials that didn’t work,” Chou says.

They discovered that changing these magnetic states modifies the material’s electronic properties, enabling low-energy operation. And unlike many other 2D materials, chromium sulfur bromide remains stable in air.

To make a transistor, the researchers pattern electrodes onto a silicon substrate, then carefully align and transfer the 2D material on top. They use tape to pick up a tiny piece of material, only a few tens of nanometers thick, and place it onto the substrate.

“A lot of researchers will use solvents or glue to do the transfer, but transistors require a very clean surface. We eliminate all those risks by simplifying this step,” Chou says.

Leveraging magnetism

This lack of contamination enables their device to outperform existing magnetic transistors. Most others can only create a weak magnetic effect, changing the flow of current by a few percent or less. Their new transistor can switch or amplify the electric current by a factor of 10.

They use an external magnetic field to change the magnetic state of the material, switching the transistor using significantly less energy than would usually be required.

The material also allows them to control the magnetic states with electric current. This is important because engineers cannot apply magnetic fields to individual transistors in an electronic device. They need to control each one electrically.

The material’s magnetic properties could also enable transistors with built-in memory, simplifying the design of logic or memory circuits.

A typical memory device has a magnetic cell to store information and a transistor to read it out. Their method can combine both into one magnetic transistor.

“Now, not only are transistors turning on and off, they are also remembering information. And because we can switch the transistor with greater magnitude, the signal is much stronger so we can read out the information faster, and in a much more reliable way,” Liu says.

Building on this demonstration, the researchers plan to further study the use of electrical current to control the device. They are also working to make their method scalable so they can fabricate arrays of transistors.

This research was supported, in part, by the Semiconductor Research Corporation, the U.S. Defense Advanced Research Projects Agency (DARPA), the U.S. National Science Foundation (NSF), the U.S. Department of Energy, the U.S. Army Research Office, and the Czech Ministry of Education, Youth, and Sports. The work was partially carried out at the MIT.nano facilities.

With global power demand from data centers expected to more than double by 2030, the MIT Energy Initiative (MITEI) in September launched an effort that brings together MIT researchers and industry experts to explore innovative solutions for powering the data-driven future. At its annual research conference, MITEI announced the Data Center Power Forum, a targeted research effort for MITEI member companies interested in addressing the challenges of data center power demand. The Data Center Power Forum builds on lessons from MITEI’s May 2025 symposium on the energy to power the expansion of artificial intelligence (AI) and focus panels related to data centers at the fall 2024 research conference.

In the United States, data centers consumed 4 percent of the country’s electricity in 2023, with demand expected to increase to 9 percent by 2030, according to the Electric Power Research Institute. Much of the growth in demand is from the increasing use of AI, which is placing an unprecedented strain on the electric grid. This surge in demand presents a serious challenge for the technology and energy sectors, government policymakers, and everyday consumers, who may see their electric bills skyrocket as a result.

“MITEI has long supported research on ways to produce more efficient and cleaner energy and to manage the electric grid. In recent years, MITEI has also funded dozens of research projects relevant to data center energy issues. Building on this history and knowledge base, MITEI’s Data Center Power Forum is convening a specialized community of industry members who have a vital stake in the sustainable growth of AI and the acceleration of solutions for powering data centers and expanding the grid,” says William H. Green, the director of MITEI and the Hoyt C. Hottel Professor of Chemical Engineering.

MITEI’s mission is to advance zero- and low-carbon solutions to expand energy access and mitigate climate change. MITEI works with companies from across the energy innovation chain, including in the infrastructure, automotive, electric power, energy, natural resources, and insurance sectors. MITEI member companies have expressed strong interest in the Data Center Power Forum and are committing to support focused research on a wide range of energy issues associated with data center expansion, Green says.

MITEI’s Data Center Power Forum will provide its member companies with reliable insights into energy supply, grid load operations and management, the built environment, and electricity market design and regulatory policy for data centers. The forum complements MIT’s deep expertise in adjacent topics such as low-power processors, efficient algorithms, task-specific AI, photonic devices, quantum computing, and the societal consequences of data center expansion. As part of the forum, MITEI’s Future Energy Systems Center is funding projects relevant to data center energy in its upcoming proposal cycles. MITEI Research Scientist Deep Deka has been named the program manager for the forum.

“Figuring out how to meet the power demands of data centers is a complicated challenge. Our research is coming at this from multiple directions, from looking at ways to expand transmission capacity within the electrical grid in order to bring power to where it is needed, to ensuring the quality of electrical service for existing users is not diminished when new data centers come online, and to shifting computing tasks to times and places when and where energy is available on the grid," said Deka.

MITEI currently sponsors substantial research related to data center energy topics across several MIT departments. The existing research portfolio includes more than a dozen projects related to data centers, including low- or zero-carbon solutions for energy supply and infrastructure, electrical grid management, and electricity market policy. MIT researchers funded through MITEI’s industry consortium are also designing more energy-efficient power electronics and processors and investigating behind-the-meter low-/no-carbon power plants and energy storage. MITEI-supported experts are studying how to use AI to optimize electrical distribution and the siting of data centers and conducting techno-economic analyses of data center power schemes. MITEI’s consortium projects are also bringing fresh perspectives to data center cooling challenges and considering policy approaches to balance the interests of shareholders. 

By drawing together industry stakeholders from across the AI and grid value chain, the Data Center Power Forum enables a richer dialog about solutions to power, grid, and carbon management problems in a noncommercial and collaborative setting.

“The opportunity to meet and to hold discussions on key data center challenges with other forum members from different sectors, as well as with MIT faculty members and research scientists, is a unique benefit of this MITEI-led effort,” Green says.

MITEI addressed the issue of data center power needs with its company members during its fall 2024 Annual Research Conference with a panel session titled, “The extreme challenge of powering data centers in a decarbonized way.” MITEI Director of Research Randall Field led a discussion with representatives from large technology companies Google and Microsoft, known as “hyperscalers,” as well as Madrid-based infrastructure developer Ferrovial S.E. and utility company Exelon Corp. Another conference session addressed the related topic, “Energy storage and grid expansion.” This past spring, MITEI focused its annual Spring Symposium on data centers, hosting faculty members and researchers from MIT and other universities, business leaders, and a representative of the Federal Energy Regulatory Commission for a full day of sessions on the topic, “AI and energy: Peril and promise.” 

A new delivery particle developed at MIT could make mRNA vaccines more effective and potentially lower the cost per vaccine dose.

In studies in mice, the researchers showed that an mRNA influenza vaccine delivered with their new lipid nanoparticle could generate the same immune response as mRNA delivered by nanoparticles made with FDA-approved materials, but at around 1/100 the dose.

“One of the challenges with mRNA vaccines is the cost,” says Daniel Anderson, a professor in MIT’s Department of Chemical Engineering and a member of MIT’s Koch Institute for Integrative Cancer Research and Institute for Medical Engineering and Science (IMES). “When you think about the cost of making a vaccine that could be distributed widely, it can really add up. Our goal has been to try to make nanoparticles that can give you a safe and effective vaccine response but at a much lower dose.”

While the researchers used their particles to deliver a flu vaccine, they could also be used for vaccines for Covid-19 and other infectious diseases, they say.

Anderson is the senior author of the study, which appears today in Nature Nanotechnology. The lead authors of the paper are Arnab Rudra, a visiting scientist at the Koch Institute; Akash Gupta, a Koch Institute research scientist; and Kaelan Reed, an MIT graduate student.

Efficient delivery

To protect mRNA vaccines from breaking down in the body after injection, they are packaged inside a lipid nanoparticle, or LNP. These fatty spheres help mRNA get into cells so that it can be translated into a fragment of a protein from a pathogen such as influenza or SARS-CoV-2.

In the new study, the MIT team sought to develop particles that can induce an effective immune response, but at a lower dose than the particles now used to deliver Covid-19 mRNA vaccines. That could not only reduce the costs per vaccine dose, but may also help to lessen the potential side effects, the researchers say.

LNPs typically consist of five elements: an ionizable lipid, cholesterol, a helper phospholipid, a polyethylene glycol lipid, and mRNA. In this study, the researchers focused on the ionizable lipid, which plays a key role in vaccine strength.

Based on their knowledge of chemical structures that might improve delivery efficiency, the researchers designed a library of new ionizable lipids. These contained cyclic structures, which can help enhance mRNA delivery, as well as chemical groups called esters, which the researchers believed could also help improve biodegradability.

The researchers then created and screened many combinations of these particle structures in mice to see which could most effectively deliver the gene for luciferase, a bioluminescent protein. Then, they took their top-performing particle and created a library of new variants, which they tested in another round of screening.

From these screens, the top LNP that emerged is one that the researchers called AMG1541. One key feature of these new LNPs is that they are more effective in dealing with a major barrier for delivery particles, known as endosomal escape. After LNPs enter cells, they are isolated in cellular compartments called endosomes, which they need to break out of to deliver their mRNA. The new particles did this more effectively than existing LNPs.

Another advantage of the new LNPs is that the ester groups in the tails make the particles degradable once they have delivered their cargo. This means they can be cleared from the body quickly, which the researchers believe could reduce side effects from the vaccine.

More powerful vaccines

To demonstrate the potential applications of the AMG1541 LNP, the researchers used it to deliver an mRNA influenza vaccine in mice. They compared this vaccine’s effectiveness to a flu vaccine made with a lipid called SM-102, which is FDA-approved and was used by Moderna in its Covid-19 vaccine.

Mice vaccinated with the new particles generated the same antibody response as mice vaccinated with the SM-102 particle, but only 1/100 of the dose was needed to generate that response, the researchers found.

“It’s almost a hundredfold lower dose, but you generate the same amount of antibodies, so that can significantly lower the dose. If it translates to humans, it should significantly lower the cost as well,” Rudra says.

Further experiments revealed that the new LNPs are better able to deliver their cargo to a critical type of immune cells called antigen-presenting cells. These cells chop up foreign antigens and display them on their surfaces, which signals other immune cells such as B and T cells to become activated against that antigen.

The new LNPs are also more likely to accumulate in the lymph nodes, where they encounter many more immune cells.

Using these particles to deliver mRNA flu vaccines could allow vaccine developers to better match the strains of flu that circulate each winter, the researchers say. “With traditional flu vaccines, they have to start being manufactured almost a year ahead of time,” Reed says. “With mRNA, you can start producing it much later in the season and get a more accurate guess of what the circulating strains are going to be, and it may help improve the efficacy of flu vaccines.”

The particles could also be adapted for vaccines for Covid-19, HIV, or any other infectious disease, the researchers say.

“We have found that they work much better than anything that has been reported so far. That’s why, for any intramuscular vaccines, we think that our LNP platforms could be used to develop vaccines for a number of diseases,” Gupta says.

The research was funded by Sanofi, the National Institutes of Health, the Marble Center for Cancer Nanomedicine, and the Koch Institute Support (core) Grant from the National Cancer Institute.

Older buildings let thousands of dollars-worth of energy go to waste each year through leaky roofs, old windows, and insufficient insulation. But even as building owners face mounting pressure to comply with stricter energy codes, making smart decisions about how to invest in efficiency is a major challenge.

Lamarr.AI, born in part from MIT research, is making the process of finding ways to improve the energy efficiency of buildings as easy as clicking a button. When customers order a building review, it triggers a coordinated symphony of drones, thermal and visible-range cameras, and artificial intelligence designed to identify problems and quantify the impact of potential upgrades. Lamarr.AI’s technology also assesses structural conditions, creates detailed 3D models of buildings, and recommends retrofits. The solution is already being used by leading organizations across facilities management as well as by architecture, engineering, and construction firms.

“We identify the root cause of the anomalies we find,” says CEO and co-founder Tarek Rakha PhD ’15. “Our platform doesn’t just say, ‘This is a hot spot and this is a cold spot.’ It specifies ‘This is infiltration or exfiltration. This is missing insulation. This is water intrusion.’ The detected anomalies are also mapped to a 3D model of the building, and there are deeper analytics, such as the cost of each retrofit and the return on investment.”

To date, the company estimates its platform has helped clients across health care, higher education, and multifamily housing avoid over $3 million in unnecessary construction and retrofit costs by recommending targeted interventions over costly full-system replacements, while improving energy performance and extending asset life. For building owners managing portfolios worth hundreds of millions of dollars, Lamarr.AI’s approach represents a fundamental shift from reactive maintenance to strategic asset management.

The founders, who also include MIT Professor John Fernández and Research Scientist Norhan Bayomi SM ’17, PhD ’21, are thrilled to see their technology accelerating the transition to more energy-efficient and higher-performing buildings.

“Reducing carbon emissions in buildings gets you the greatest return on investment in terms of climate interventions, but what has been needed are the technologies and tools to help the real estate and construction sectors make the right decisions in a timely and economical way,” Fernández says.

Automating building scans

Bayomi and Rakha completed their PhDs in the MIT Department of Architecture’s Building Technology Program. For her thesis, Bayomi developed technology to detect features of building exteriors and classify thermal anomalies through scans of buildings, with a specific focus on the impact of heat waves on low-income communities. Bayomi and her collaborators eventually deployed the system to detect air leaks as part of a partnership with a community in New York City.

After graduating MIT, Rakha became an assistant professor at Syracuse University. In 2015, together with fellow Syracuse University Professor Senem Velipasalar, he began developing his concept for drone-based building analytics — an idea that later received support through a grant from New York State’s Department of Economic Development. In 2019, Bayomi and Fernández joined the project, and the team received a $1.8 million research award from the U.S. Department of Energy.

“The technology is like giving a building an MRI using drones, infrared imaging, visible light imaging, and proprietary AI that we developed through computer vision technology, along with large language models for report generation,” Rakha explains.

“When we started the research, we saw firsthand how vulnerable communities were suffering from inefficient buildings, but couldn’t afford comprehensive diagnostics,” Bayomi says. “We knew that if we could automate this process and reduce costs while improving accuracy, we’d unlock a massive market. Now we’re seeing demand from everyone, from municipal buildings to major institutional portfolios.”

Lamarr.AI was officially founded in 2021 to commercialize the technology, and the founders wasted no time tapping into MIT’s entrepreneurial ecosystem. First, they received a small seed grant from the MIT Sandbox Innovation Fund. In 2022, they won the MITdesignX prize and were semifinalists in the MIT $100K Entrepreneurship Competition. The founders named the company after Hedy Lamarr, the famous actress and inventor of a patented technology that became the basis for many modern secure communications.

Current methods for detecting air leaks in buildings utilize fan pressurizers or smoke. Contractors or building engineers may also spot-check buildings with handheld infrared cameras to manually identify temperature differences across individual walls, windows, and ductwork.

Lamarr.AI’s system can perform building inspections far more quickly. Building managers can order the company’s scans online and select when they’d like the drone to fly. Lamarr.AI partners with drone companies worldwide to fly off-the-shelf drones around buildings, providing them with flight plans and specifications for success. Images are then uploaded onto Lamarr.AI’s platform for automated analysis.

“As an example, a survey of a 180,000-square-foot building like the MIT Schwarzman College of Computing, which we scanned, produces around 2,000 images,” Fernández says. “For someone to go through those manually would take a couple of weeks. Our models autonomously analyze those images in a few seconds.”

After the analysis, Lamarr.AI’s platform generates a report that includes the suspected root cause of every weak point found, an estimated cost to correct that problem, and its estimated return on investment using advanced building energy simulations.

“We knew if we were able to quickly, inexpensively, and accurately survey the thermal envelope of buildings and understand their performance, we would be addressing a huge need in the real estate, building construction, and built environment sectors,” Fernández explains. “Thermal anomalies are a huge cause of unwanted heat loss, and more than 45 percent of construction defects are tied to envelope failures.”

The ability to operate at scale is especially attractive to building owners and operators, who often manage large portfolios of buildings across multiple campuses.

“We see Lamarr.AI becoming the premier solution for building portfolio diagnostics and prognosis across the globe, where every building can be equipped not just for the climate crisis, but also to minimize energy losses and be more efficient, safer, and sustainable,” Rakha says.

Building science for everyone

Lamarr.AI has worked with building operators across the U.S. as well as in Canada, the United Kingdom, and the United Arab Emirates.

In June, Lamarr.AI partnered with the City of Detroit, with support from Newlab and Michigan Central, to inspect three municipal buildings to identify areas for improvement. Across two of the buildings, the system identified more than 460 problems like insulation gaps and water leaks. The findings were presented in a report that also utilized energy simulations to demonstrate that upgrades, such as window replacements and targeted weatherization, could reduce HVAC energy use by up to 22 percent.

The entire process took a few days. The founders note that it was the first building inspection drone flight to utilize an off-site operator, an approach that further enhances the scalability of their platform. It also helps further reduce costs, which could make building scans available to a broader swath of people around the world.

“We’re democratizing access to very high-value building science expertise that previously cost tens of thousands per audit,” Bayomi says. “Our platform makes advanced diagnostics affordable enough for routine use, not just one-time assessments. The bigger vision is automated, regular building health monitoring that keeps facilities teams informed in real-time, enabling proactive decisions rather than reactive crisis management. When building intelligence becomes continuous and accessible, operators can optimize performance systematically rather than waiting for problems to emerge.”

Adoption of new tools and technologies occurs when users largely perceive them as reliable, accessible, and an improvement over the available methods and workflows for the cost. Five PhD students from the inaugural class of the MIT-IBM Watson AI Lab Summer Program are utilizing state-of-the-art resources, alleviating AI pain points, and creating new features and capabilities to promote AI usefulness and deployment — from learning when to trust a model that predicts another’s accuracy to more effectively reasoning over knowledge bases. Together, the efforts from the students and their mentors form a through-line, where practical and technically rigorous research leads to more dependable and valuable models across domains.

Building probes, routers, new attention mechanisms, synthetic datasets, and program-synthesis pipelines, the students’ work spans safety, inference efficiency, multimodal data, and knowledge-grounded reasoning. Their techniques emphasize scaling and integration, with impact always in sight.

Learning to trust, and when

MIT math graduate student Andrey Bryutkin’s research prioritizes the trustworthiness of models. He seeks out internal structures within problems, such as equations governing a system and conservation laws, to understand how to leverage them to produce more dependable and robust solutions. Armed with this and working with the lab, Bryutkin developed a method to peer into the nature of large learning models (LLMs) behaviors. Together with the lab’s Veronika Thost of IBM Research and Marzyeh Ghassemi — associate professor and the Germeshausen Career Development Professor in the MIT Department of Electrical Engineering and Computer Science (EECS) and a member of the Institute of Medical Engineering Sciences and the Laboratory for Information and Decision Systems — Bryutkin explored the “uncertainty of uncertainty” of LLMs. 

Classically, tiny feed-forward neural networks two-to-three layers deep, called probes, are trained alongside LLMs and employed to flag untrustworthy answers from the larger model to developers; however, these classifiers can also produce false negatives and only provide point estimates, which don’t offer much information about when the LLM is failing. Investigating safe/unsafe prompts and question-answer tasks, the MIT-IBM team used prompt-label pairs, as well as the hidden states like activation vectors and last tokens from an LLM, to measure gradient scores, sensitivity to prompts, and out-of-distribution data to determine how reliable the probe was and learn areas of data that are difficult to predict. Their method also helps identify potential labeling noise. This is a critical function, as the trustworthiness of AI systems depends entirely on the quality and accuracy of the labeled data they are built upon. More accurate and consistent probes are especially important for domains with critical data in applications like IBM’s Granite Guardian family of models.

Another way to ensure trustworthy responses to queries from an LLM is to augment them with external, trusted knowledge bases to eliminate hallucinations. For structured data, such as social media connections, financial transactions, or corporate databases, knowledge graphs (KG) are natural fits; however, communications between the LLM and KGs often use fixed, multi-agent pipelines that are computationally inefficient and expensive. Addressing this, physics graduate student Jinyeop Song, along with lab researchers Yada Zhu of IBM Research and EECS Associate Professor Julian Shun created a single-agent, multi-turn, reinforcement learning framework that streamlines this process. Here, the group designed an API server hosting Freebase and Wikidata KGs, which consist of general web-based knowledge data, and a LLM agent that issues targeted retrieval actions to fetch pertinent information from the server. Then, through continuous back-and-forth, the agent appends the gathered data from the KGs to the context and responds to the query. Crucially, the system uses reinforcement learning to train itself to deliver answers that strike a balance between accuracy and completeness. The framework pairs an API server with a single reinforcement learning agent to orchestrate data-grounded reasoning with improved accuracy, transparency, efficiency, and transferability.

Spending computation wisely

The timeliness and completeness of a model’s response carry similar weight to the importance of its accuracy. This is especially true for handling long input texts and those where elements, like the subject of a story, evolve over time, so EECS graduate student Songlin Yang is re-engineering what models can handle at each step of inference. Focusing on transformer limitations, like those in LLMs, the lab’s Rameswar Panda of IBM Research and Yoon Kim, the NBX Professor and associate professor in EECS, joined Yang to develop next-generation language model architectures beyond transformers.

Transformers face two key limitations: high computational complexity in long-sequence modeling due to the softmax attention mechanism, and limited expressivity resulting from the weak inductive bias of RoPE (rotary positional encoding). This means that as the input length doubles, the computational cost quadruples. RoPE allows transformers to understand the sequence order of tokens (i.e., words); however, it does not do a good job capturing internal state changes over time, like variable values, and is limited to the sequence lengths seen during training.

To address this, the MIT-IBM team explored theoretically grounded yet hardware-efficient algorithms. As an alternative to softmax attention, they adopted linear attention, reducing the quadratic complexity that limits the feasible sequence length. They also investigated hybrid architectures that combine softmax and linear attention to strike a better balance between computational efficiency and performance.

Increasing expressivity, they replaced RoPE with a dynamic reflective positional encoding based on the Householder transform. This approach enables richer positional interactions for deeper understanding of sequential information, while maintaining fast and efficient computation. The MIT-IBM team’s advancement reduces the need for transformers to break problems into many steps, instead enabling them to handle more complex subproblems with fewer inference tokens.

Visions anew

Visual data contain multitudes that the human brain can quickly parse, internalize, and then imitate. Using vision-language models (VLMs), two graduate students are exploring ways to do this through code.

Over the past two summers and under the advisement of Aude Oliva, MIT director of the MIT-IBM Watson AI Lab and a senior research scientist in the Computer Science and Artificial Intelligence Laboratory; and IBM Research’s Rogerio Feris, Dan Gutfreund, and Leonid Karlinsky (now at Xero), Jovana Kondic of EECS has explored visual document understanding, specifically charts. These contain elements, such as data points, legends, and axes labels, that require optical character recognition and numerical reasoning, which models still struggle with. In order to facilitate the performance on tasks such as these, Kondic’s group set out to create a large, open-source, synthetic chart dataset from code that could be used for training and benchmarking. 

With their prototype, ChartGen, the researchers created a pipeline that passes seed chart images through a VLM, which is prompted to read the chart and generate a Python script that was likely used to create the chart in the first place. The LLM component of the framework then iteratively augments the code from many charts to ultimately produce over 200,000 unique pairs of charts and their codes, spanning nearly 30 chart types, as well as supporting data and annotation like descriptions and question-answer pairs about the charts. The team is further expanding their dataset, helping to enable critical multimodal understanding to data visualizations for enterprise applications like financial and scientific reports, blogs, and more.

Instead of charts, EECS graduate student Leonardo Hernandez Cano has his eyes on digital design, specifically visual texture generation for CAD applications and the goal of discovering efficient ways to enable to capabilities in VLMs. Teaming up with the lab groups led by Armando Solar-Lezama, EECS professor and Distinguished Professor of Computing in the MIT Schwarzman College of Computing, and IBM Research’s Nathan Fulton, Hernandez Cano created a program synthesis system that learns to refine code on its own. The system starts with a texture description given by a user in the form of an image. It then generates an initial Python program, which produces visual textures, and iteratively refines the code with the goal of finding a program that produces a texture that matches the target description, learning to search for new programs from the data that the system itself produces. Through these refinements, the novel program can create visualizations with the desired luminosity, color, iridescence, etc., mimicking real materials.

When viewed together, these projects, and the people behind them, are making a cohesive push toward more robust and practical artificial intelligence. By tackling the core challenges of reliability, efficiency, and multimodal reasoning, the work paves the way for AI systems that are not only more powerful, but also more dependable and cost-effective, for real-world enterprise and scientific applications.

The MIT Living Climate Futures Lab (LCFL) centers the human dimensions of climate change, bringing together expertise from across MIT to address one of the world’s biggest challenges.

The LCFL has three main goals: “addressing how climate change plays out in everyday life, focusing on community-oriented partnerships, and encouraging cross-disciplinary conversations around climate change on campus,” says Chris Walley, the SHASS Dean’s Distinguished Professor of Anthropology and head of MIT’s Anthropology Section. “We think this is a crucial direction for MIT and will make a strong statement about the kind of human-centered, interdisciplinary work needed to tackle this issue.”

Walley is faculty lead of LCFL, working in collaboration with a group of 19 faculty colleagues and researchers. The LCFL began to coalesce in 2022 when MIT faculty and affiliates already working with communities dealing with climate change issues organized a symposium, inviting urban farmers, place-based environmental groups, and others to MIT. Since then, the lab has consolidated the efforts of faculty and affiliates representing disciplines from across the MIT School of Humanities, Arts, and Social Sciences (SHASS) and the Institute.

Amah Edoh, a cultural anthropologist and managing director of LCFL, says the lab’s collaboration with community organizations and development of experiential learning classes aims to bridge the gap that can exist between the classroom and the real world.

“Sometimes we can find ourselves in a bubble where we’re only in conversation with other people from within academia or our own field of practice. There can be a disconnect between what students are learning somewhat abstractly and the ‘real world’ experience of the issues” Edoh says. “By taking up topics from the multidimensional approach that experiential learning makes possible, students learn to take complexity as a given, which can help to foster more critical thinking in them, and inform their future practice in profound ways.”

Edoh points out that the effects of climate change play out in a huge array of areas: health, food security, livelihoods, housing, and governance structures, to name a few.

“The Living Climate Futures Lab supports MIT researchers in developing the long-term collaborations with community partners that are essential to adequately identifying and responding to the challenges that climate change creates in everyday life,” she says.

Manduhai Buyandelger, professor of anthropology and one of the participants in LCFL, developed the class 21A.S01 (Anthro-Engineering: Decarbonization at the Million-Person Scale), which has in turn sparked related classes. The goal is “to merge technological innovation with people-centered environments.” Working closely with residents of Ulaanbaatar, Mongolia, Buyandelger and collaborator Mike Short, the Class of 1941 Professor of Nuclear Science and Engineering, helped develop a molten salt heat bank as a reusable energy source.

“My work with Mike Short on energy and alternative heating in Mongolia helps to cultivate a new generation of creative and socially minded engineers who prioritize people in thinking about technical solutions,” Buyandelger says, adding, “In our course, we collaborate on creating interdisciplinary methods where we fuse anthropological methods with engineering innovations so that we can expand and deepen our approach to mitigate climate change.”

Iselle Barrios ’25, says 21A.S01 was her first anthropology course. She traveled to Mongolia and was able to experience firsthand all the ways in which the air pollution and heating problem was much larger and more complicated than it seemed from MIT’s Cambridge, Massachusetts, campus.

“It was my first exposure to anthropological and STS critiques of science and engineering, as well as international development,” says Barrios, a chemical engineering major. “It fundamentally reshaped the way I see the role of technology and engineers in the broader social context in which they operate. It really helped me learn to think about problems in a more holistic and people-centered way.”

LCFL participant Alvin Harvey, a postdoc in the MIT Media Lab’s Space Enabled Research Group and a citizen of the Navajo Nation, works to incorporate traditional knowledge in engineering and science to “support global stewardship of earth and space ecologies.”

"I envision the Living Climate Futures Lab as a collaborative space that can be an igniter and sustainer of relationships, especially between MIT and those whose have generational and cultural ties to land and space that is being impacted by climate change,” Harvey says. “I think everyone in our lab understands that protecting our climate future is a collective journey."

Kate Brown, the Thomas M. Siebel Distinguished Professor in History of Science, is also a participant in LCFL. Her current interest is urban food sovereignty movements, in which working-class city dwellers used waste to create “the most productive agriculture in recorded human history,” Brown says. While pursuing that work, Brown has developed relationships and worked with urban farmers in Mansfield, Ohio, as well as in Washington and Amsterdam.

Brown and Susan Solomon, the Lee and Geraldine Martin Professor of Environmental Studies and Chemistry, teach a class called STS.055 (Living Dangerously: Environmental Programs from 1900 to Today) that presents the environmental problems and solutions of the 20th century, and how some “solutions” created more problems over time. Brown also plans to teach a class on the history of global food production once she gets access to a small plot of land on campus for a lab site.

“The Living Climate Futures Lab gives us the structure and flexibility to work with communities that are struggling to find solutions to the problems being created by the climate crisis,” says Brown.

Earlier this year, the MIT Human Insight Collaborative (MITHIC) selected the Living Climate Futures Lab as its inaugural Faculty-Driven Initiative (FDI), which comes with a $500,000 seed grant.

MIT Provost Anantha Chandrakasan, co-chair of MITHIC, says the LCFL exemplifies how we can confront the climate crisis by working in true partnership with the communities most affected.

“By combining scientific insight with cultural understanding and lived experience, this initiative brings a deeper dimension to MIT’s climate efforts — one grounded in collaboration, empathy, and real-world impact,” says Chandrakasan.

Agustín Rayo, the Kenan Sahin Dean of SHASS and co-chair of MITHIC, says the LCFL is precisely the type of interdisciplinary collaboration the FDI program was designed to support.

"By bringing together expertise from across MIT, I am confident the Living Climate Futures Lab will make significant contributions in the Institute’s effort to address the climate crisis," says Rayo.

Walley said the seed grant will support a second symposium in 2026 to be co-designed with community groups, a suite of experiential learning classes, workshops, a speaker series, and other programming. Throughout this development phase, the lab will solicit donor support to build it into an ongoing MIT initiative and a leader in the response to climate change.

Superconductors are like the express trains in a metro system. Any electricity that “boards” a superconducting material can zip through it without stopping and losing energy along the way. As such, superconductors are extremely energy efficient, and are used today to power a variety of applications, from MRI machines to particle accelerators.

But these “conventional” superconductors are somewhat limited in terms of uses because they must be brought down to ultra-low temperatures using elaborate cooling systems to keep them in their superconducting state. If superconductors could work at higher, room-like temperatures, they would enable a new world of technologies, from zero-energy-loss power cables and electricity grids to practical quantum computing systems. And so scientists at MIT and elsewhere are studying “unconventional” superconductors — materials that exhibit superconductivity in ways that are different from, and potentially more promising than, today’s superconductors.

In a promising breakthrough, MIT physicists have today reported their observation of new key evidence of unconventional superconductivity in “magic-angle” twisted tri-layer graphene (MATTG) — a material that is made by stacking three atomically-thin sheets of graphene at a specific angle, or twist, that then allows exotic properties to emerge.

MATTG has shown indirect hints of unconventional superconductivity and other strange electronic behavior in the past. The new discovery, reported in the journal Science, offers the most direct confirmation yet that the material exhibits unconventional superconductivity.

In particular, the team was able to measure MATTG’s superconducting gap — a property that describes how resilient a material’s superconducting state is at given temperatures. They found that MATTG’s superconducting gap looks very different from that of the typical superconductor, meaning that the mechanism by which the material becomes superconductive must also be different, and unconventional.

“There are many different mechanisms that can lead to superconductivity in materials,” says study co-lead author Shuwen Sun, a graduate student in MIT’s Department of Physics. “The superconducting gap gives us a clue to what kind of mechanism can lead to things like room-temperature superconductors that will eventually benefit human society.”

The researchers made their discovery using a new experimental platform that allows them to essentially “watch” the superconducting gap, as the superconductivity emerges in two-dimensional materials, in real-time. They plan to apply the platform to further probe MATTG, and to map the superconducting gap in other 2D materials — an effort that could reveal promising candidates for future technologies.

“Understanding one unconventional superconductor very well may trigger our understanding of the rest,” says Pablo Jarillo-Herrero, the Cecil and Ida Green Professor of Physics at MIT and the senior author of the study. “This understanding may guide the design of superconductors that work at room temperature, for example, which is sort of the Holy Grail of the entire field.”

The study’s other co-lead author is Jeong Min Park PhD ’24; Kenji Watanabe and Takashi Taniguchi of the National Institute for Materials Science in Japan are also co-authors.

The ties that bind

Graphene is a material that comprises a single layer of carbon atoms that are linked in a hexagonal pattern resembling chicken wire. A sheet of graphene can be isolated by carefully exfoliating an atom-thin flake from a block of graphite (the same stuff of pencil lead). In the 2010s, theorists predicted that if two graphene layers were stacked at a very special angle, the resulting structure should be capable of exotic electronic behavior.

In 2018, Jarillo-Herrero and his colleagues became the first to produce magic-angle graphene in experiments, and to observe some of its extraordinary properties. That discovery sprouted an entire new field known as “twistronics,” and the study of atomically thin, precisely twisted materials. Jarillo-Herrero’s group has since studied other configurations of magic-angle graphene with two, three, and more layers, as well as stacked and twisted structures of other two-dimensional materials. Their work, along with other groups, have revealed some signatures of unconventional superconductivity in some structures.

Superconductivity is a state that a material can exhibit under certain conditions (usually at very low temperatures). When a material is a superconductor, any electrons that pass through can pair up, rather than repelling and scattering away. When they couple up in what is known as “Cooper pairs,” the electrons can glide through a material without friction, instead of knocking against each other and flying away as lost energy. This pairing up of electrons is what enables superconductivity, though the way in which they are bound can vary.

“In conventional superconductors, the electrons in these pairs are very far away from each other, and weakly bound,” says Park. “But in magic-angle graphene, we could already see signatures that these pairs are very tightly bound, almost like a molecule. There were hints that there is something very different about this material.”

Tunneling through

In their new study, Jarillo-Herrero and his colleagues aimed to directly observe and confirm unconventional superconductivity in a magic-angle graphene structure. To do so, they would have to measure the material’s superconducting gap.

“When a material becomes superconducting, electrons move together as pairs rather than individually, and there’s an energy ‘gap’ that reflects how they’re bound,” Park explains. “The shape and symmetry of that gap tells us the underlying nature of the superconductivity.”

Scientists have measured the superconducting gap in materials using specialized techniques, such as tunneling spectroscopy. The technique takes advantage of a quantum mechanical property known as “tunneling.” At the quantum scale, an electron behaves not just as a particle, but also as a wave; as such, its wave-like properties enable an electron to travel, or “tunnel,” through a material, as if it could move through walls.

Such tunneling spectroscopy measurements can give an idea of how easy it is for an electron to tunnel into a material, and in some sense, how tightly packed and bound the electrons in the material are. When performed in a superconducting state, it can reflect the properties of the superconducting gap. However, tunneling spectroscopy alone cannot always tell whether the material is, in fact, in a superconducting state. Directly linking a tunneling signal to a genuine superconducting gap is both essential and experimentally challenging.

In their new work, Park and her colleagues developed an experimental platform that combines electron tunneling with electrical transport — a technique that is used to gauge a material’s superconductivity, by sending current through and continuously measuring its electrical resistance (zero resistance signals that a material is in a superconducting state).

The team applied the new platform to measure the superconducting gap in MATTG. By combining tunneling and transport measurements in the same device, they could unambiguously identify the superconducting tunneling gap, one that appeared only when the material exhibited zero electrical resistance, which is the hallmark of superconductivity. They then tracked how this gap evolved under varying temperature and magnetic fields. Remarkably, the gap displayed a distinct V-shaped profile, which was clearly different from the flat and uniform shape of conventional superconductors.

This V shape reflects a certain unconventional mechanism by which electrons in MATTG pair up to superconduct. Exactly what that mechanism is remains unknown. But the fact that the shape of the superconducting gap in MATTG stands out from that of the typical superconductor provides key evidence that the material is an unconventional superconductor.

In conventional superconductors, electrons pair up through vibrations of the surrounding atomic lattice, which effectively jostle the particles together. But Park suspects that a different mechanism could be at work in MATTG.

“In this magic-angle graphene system, there are theories explaining that the pairing likely arises from strong electronic interactions rather than lattice vibrations,” she posits. “That means electrons themselves help each other pair up, forming a superconducting state with special symmetry.”

Going forward, the team will test other two-dimensional twisted structures and materials using the new experimental platform.

“This allows us to both identify and study the underlying electronic structures of superconductivity and other quantum phases as they happen, within the same sample,” Park says. “This direct view can reveal how electrons pair and compete with other states, paving the way to design and control new superconductors and quantum materials that could one day power more efficient technologies or quantum computers.”

This research was supported, in part, by the U.S. Army Research Office, the U.S. Air Force Office of Scientific Research, the MIT/MTL Samsung Semiconductor Research Fund, the Sagol WIS-MIT Bridge Program, the National Science Foundation, the Gordon and Betty Moore Foundation, and the Ramon Areces Foundation.

Traditional folk ballads are one of our most enduring forms of cultural expression. They can also be lost to society, forgotten over time. That’s why, in the mid-1700s, when a Scottish woman named Anna Gordon was found to know three dozen ancient ballads, collectors tried to document all of these songs — a volume of work that became a kind of sensation in its time, a celebrated piece of cultural heritage.

That story is told in MIT Professor Emerita Ruth Perry’s latest book, “The Ballad World of Anna Gordon, Mrs. Brown of Falkland,” published this year by Oxford University Press. In it, Perry details what we know about the ways folk ballads were created and transmitted; how Anna Gordon came to know so many; the social and political climate in which they existed; and why these songs meant so much in Scotland and elsewhere in the Atlantic world. Indeed, Scottish immigrants brought their music to the U.S., among other places.

MIT News sat down with Perry, who is MIT’s Ann Fetter Friedlaender Professor of Humanities, Emerita, to talk about the book.

Q: This is fascinating topic with a lot of threads woven together. To you, what is the book about?

A: It’s really three books. It’s a book about Anna Gordon and her family, a very interesting middle-class family living in Aberdeen in the middle of the 18th century. And it’s a book about balladry and what a ballad is — a story told in song, and ballads are the oldest known poetry in English. Some of them are gorgeous. Third, it’s a book about the relationship between Scotland and England, the effects of the Jacobite uprising in 1745, social attitudes, how people lived, what they ate, education — it’s very much about 18th century Scotland.

Q: Okay, who was Anna Gordon, and what was her family milieu?

A: Anna’s father, Thomas Gordon, was a professor at King’s College, now the University of Aberdeen. He was a professor of humanity, which in those days meant Greek and Latin, and was well-connected to the intellectual community of the Scottish Enlightenment. A friend of his, an Edinburgh writer, lawyer, and judge, William Tytler, who heard cases all over the country and always stayed with Thomas Gordon and his family when he came to Aberdeen, was intensely interested in Scottish traditional music. He found out that Anna Gordon had learned all these ballads as a child, from her mother and aunt and some servants. Tytler asked if she would write them down, both tunes and words.

That was the earliest manuscript of ballads ever collected from a named person in Scotland. Once it was in existence, all kinds of people wanted to see it; it got spread throughout the country. In my book, I detail much of the excitement over this manuscript.

The thing about Anna’s ballads is: It’s not just that there are more of them, and more complete versions that are fuller, with more verses. They’re more beautiful. The language is more archaic, and there are marvelous touches. It is thought, and I agree, that Anna Gordon was an oral poet. As she remembered ballads and reproduced them, she improved on them. She had a great memory for the best bits and would improve other parts.

Q: How did it come about that at this time, a woman such as Anna Gordon would be the keeper and creator of cultural knowledge?

A: Women were more literate in Scotland than elsewhere. The Scottish Parliament passed an act in 1695 requiring every parish in the Church of Scotland to have not only a minister, but a teacher. Scotland was the most literate country in Europe in the 18th century. And those parish schoolmasters taught local kids. The parents did have to pay a few pennies for their classes, and, true, more parents paid for sons than for daughters. But there were daughters who took classes. And there were no opportunities like this in England at the time. Education was better for women in Scotland. So was their legal position, under common law in Scotland. When the Act of Union was formed in 1707, Scotland retained its own legal system, which had more extensive rights for women than in England.

Q: I know it’s complex, but generally, why was this?

A: Scotland was a much more democratic country, culture, and society than England, period. When Elizabeth I died in 1603, the person who inherited the throne was the King of Scotland James VI, who went to England with his court — which included the Scottish aristocracy. So, the Scottish aristocracy ended up in London. I’m sure they went back to their hunting lodges for the hunting season, but they didn’t live there [in Scotland] and they didn’t set the tone of the country. It was democratized because all that was left were a lot of lawyers and ministers and teachers.

Q: What is distinctive about the ballads in this corpus of songs Anna Gordon knew and documented?

A: A common word about ballads is that there’s a high body count, and they’re all about people dying and killing each other. But that is not true of Anna Gordon’s ballads. They’re about younger women triumphing in the world, often against older women, which is interesting, and even more often against fathers. The ballads are about family discord, inheritance, love, fidelity, lack of fidelity, betrayal. There are ballads about fighting and bloodshed, but not so many. They’re about the human condition. And they have interesting qualities because they’re oral poetry, composed and remembered and changed and transmitted from mouth to ear and not written down. There are repetitions and parallelisms, and other hallmarks of oral poetry. The sort of thing you learned when you read Homer.

Q: So is this a form of culture generated in opposition to those controlling society? Or at least, one that’s popular regardless of what some elites thought?

A: It is in Scotland, because of the enmity between Scotland and England. We’re talking about the period of Great Britain when England is trying to gobble up Scotland and some Scottish folks don’t want that. They want to retain their Scottishness. And the ballad was a Scottish tradition that was not influenced by England. That’s one reason balladry was so important in 18th-century Scotland. Everybody was into balladry partly because it was a unique part of Scottish culture.

Q: To that point, it seems like an unexpected convergence, for the time, to see a more middle-class woman like Anna Gordon transmitting ballads that had often been created and sung by people of all classes.

A: Yes. At first I thought I was just working on a biography of Anna Gordon. But it’s fascinating how the culture was transmitted, how intellectually rich that society was, how much there is to examine in Scottish culture and society of the 18th century. Today people may watch “Outlander,” but they still wouldn’t know anything about this!

A new 3D human brain tissue platform developed by MIT researchers is the first to integrate all major brain cell types, including neurons, glial cells, and the vasculature, into a single culture. 

Grown from individual donors’ induced pluripotent stem cells, these models — dubbed Multicellular Integrated Brains (miBrains) — replicate key features and functions of human brain tissue, are readily customizable through gene editing, and can be produced in quantities that support large-scale research.

Although each unit is smaller than a dime, miBrains may be worth a great deal to researchers and drug developers who need more complex living lab models to better understand brain biology and treat diseases.

“The miBrain is the only in vitro system that contains all six major cell types that are present in the human brain,” says Li-Huei Tsai, Picower Professor, director of The Picower Institute for Learning and Memory, and a senior author of the open-access study describing miBrains, published Oct. 17 in the Proceedings of the National Academy of Sciences.

“In their first application, miBrains enabled us to discover how one of the most common genetic markers for Alzheimer’s disease alters cells’ interactions to produce pathology,” she adds.

Tsai’s co-senior authors are Robert Langer, David H. Koch (1962) Institute Professor, and Joel Blanchard, associate professor in the Icahn School of Medicine at Mt. Sinai in New York, and a former Tsai Laboratory postdoc. The study is led by Alice Stanton, former postdoc in the Langer and Tsai labs and now assistant professor at Harvard Medical School and Massachusetts General Hospital, and Adele Bubnys, a former Tsai lab postdoc and current senior scientist at Arbor Biotechnologies.

Benefits from two kinds of models

The more closely a model recapitulates the brain’s complexity, the better suited it is for extrapolating how human biology works and how potential therapies may affect patients. In the brain, neurons interact with each other and with various helper cells, all of which are arranged in a three-dimensional tissue environment that includes blood vessels and other components. All of these interactions are necessary for health, and any of them can contribute to disease.

Simple cultures of just one or a few cell types can be created in quantity relatively easily and quickly, but they cannot tell researchers about the myriad interactions that are essential to understanding health or disease. Animal models embody the brain’s complexity, but can be difficult and expensive to maintain, slow to yield results, and different enough from humans to yield occasionally divergent results.

MiBrains combine advantages from each type of model, retaining much of the accessibility and speed of lab-cultured cell lines while allowing researchers to obtain results that more closely reflect the complex biology of human brain tissue. Moreover, they are derived from individual patients, making them personalized to an individual’s genome. In the model, the six cell types self-assemble into functioning units, including blood vessels, immune defenses, and nerve signal conduction, among other features. Researchers ensured that miBrains also possess a blood-brain-barrier capable of gatekeeping which substances may enter the brain, including most traditional drugs.

“The miBrain is very exciting as a scientific achievement,” says Langer. “Recent trends toward minimizing the use of animal models in drug development could make systems like this one increasingly important tools for discovering and developing new human drug targets.”

Two ideal blends for functional brain models

Designing a model integrating so many cell types presented challenges that required many years to overcome. Among the most crucial was identifying a substrate able to provide physical structure for cells and support their viability. The research team drew inspiration from the environment that surrounds cells in natural tissue, the extracellular matrix (ECM). The miBrain’s hydrogel-based “neuromatrix” mimics the brain’s ECM with a custom blend of polysaccharides, proteoglycans, and basement membrane that provide a scaffold for all the brain’s major cell types while promoting the development of functional neurons.

A second blend would also prove critical: the proportion of cells that would result in functional neurovascular units. The actual ratios of cell types have been a matter of debate for the last several decades, with even the more advanced methodologies providing only rough brushstrokes for guidance, for example 45-75 percent for oligodendroglia of all cells or 19-40 percent for astrocytes.

The researchers developed the six cell types from patient-donated induced pluripotent stem cells, verifying that each cultured cell type closely recreated naturally-occurring brain cells. Then, the team experimentally iterated until they hit on a balance of cell types that resulted in functional, properly structured neurovascular units. This laborious process would turn out to be an advantageous feature of miBrains: because cell types are cultured separately, they can each be genetically edited so that the resulting model is tailored to replicate specific health and disease states.

“Its highly modular design sets the miBrain apart, offering precise control over cellular inputs, genetic backgrounds, and sensors — useful features for applications such as disease modeling and drug testing,” says Stanton.

Alzheimer’s discovery using miBrain

To test miBrain’s capabilities, the researchers embarked on a study of the gene variant APOE4, which is the strongest genetic predictor for the development of Alzheimer’s disease. Although one brain cell type, astrocytes, are known to be a primary producer of the APOE protein, the role that astrocytes carrying the APOE4 variant play in disease pathology is poorly understood.

MiBrains were well-suited to the task for two reasons. First of all, they integrate astrocytes with the brain’s other cell types, so that their natural interactions with other cells can be mimicked. Second, because the platform allowed the team to integrate cell types individually, APOE4 astrocytes could be studied in cultures where all other cell types carried APOE3, a gene variant that does not increase Alzheimer’s risk. This enabled the researchers to isolate the contribution APOE4 astrocytes make to pathology.

In one experiment, the researchers examined APOE4 astrocytes cultured alone, versus ones in APOE4 miBrains. They found that only in the miBrains did the astrocytes express many measures of immune reactivity associated with Alzheimer’s disease, suggesting the multicellular environment contributes to that state.

The researchers also tracked the Alzheimer’s-associated proteins amyloid and phosphorylated tau, and found all-APOE4 miBrains accumulated them, whereas all-APOE3 miBrains did not, as expected. However, in APOE3 miBrains with APOE4 astrocytes, they found that APOE4 miBrains still exhibited amyloid and tau accumulation.

Then the team dug deeper into how APOE4 astrocytes’ interactions with other cell types might lead to their contribution to disease pathology. Prior studies have implicated molecular cross-talk with the brain’s microglia immune cells. Notably, when the researchers cultured APOE4 miBrains without microglia, their production of phosphorylated tau was significantly reduced. When the researchers dosed APOE4 miBrains with culture media from astrocytes and microglia combined, phosphorylated tau increased, whereas when they dosed them with media from cultures of astrocytes or microglia alone, the tau production did not increase. The results therefore provided new evidence that molecular cross-talk between microglia and astrocytes is indeed required for phosphorylated tau pathology.

In the future, the research team plans to add new features to miBrains to more closely model characteristics of working brains, such as leveraging microfluidics to add flow through blood vessels, or single-cell RNA sequencing methods to improve profiling of neurons.

Researchers expect that miBrains could advance research discoveries and treatment modalities for Alzheimer’s disease and beyond. 

“Given its sophistication and modularity, there are limitless future directions,” says Stanton. “Among them, we would like to harness it to gain new insights into disease targets, advanced readouts of therapeutic efficacy, and optimization of drug delivery vehicles.”

“I’m most excited by the possibility to create individualized miBrains for different individuals,” adds Tsai. “This promises to pave the way for developing personalized medicine.”

Funding for the study came from the BT Charitable Foundation, Freedom Together Foundation, the Robert A. and Renee E. Belfer Family, Lester A. Gimpelson, Eduardo Eurnekian, Kathleen and Miguel Octavio, David B. Emmes, the Halis Family, the Picower Institute, and an anonymous donor.

A large-scale screen of tuberculosis proteins has revealed several possible antigens that could be developed as a new vaccine for TB, the world’s deadliest infectious disease.

In the new study, a team of MIT biological engineers was able to identify a handful of immunogenic peptides, out of more than 4,000 bacterial proteins, that appear to stimulate a strong response from a type of T cells responsible for orchestrating immune cells’ response to infection.

There is currently only one vaccine for tuberculosis, known as BCG, which is a weakened version of a bacterium that causes TB in cows. This vaccine is widely administered in some parts of the world, but it poorly protects adults against pulmonary TB. Worldwide, tuberculosis kills more than 1 million people every year.

“There’s still a huge TB burden globally that we’d like to make an impact on,” says Bryan Bryson, an associate professor of biological engineering at MIT and a member of the Ragon Institute of Mass General Brigham, MIT, and Harvard. “What we’ve tried to do in this initial TB vaccine is focus on antigens that we saw frequently in our screen and also appear to stimulate a response in T cells from people with prior TB infection.”

Bryson and Forest White, the Ned C. and Janet C. Rice Professor of Biological Engineering at MIT, and a member of the Koch Institute for Integrative Cancer Research, are the senior authors of the study, which appears today in Science Translational Medicine. Owen Leddy PhD ’25 is the paper’s lead author.

Identifying vaccine targets

Since the BCG vaccine was developed more than 100 years ago, no other TB vaccines have been approved for use. Mycobacterium tuberculosis produces more than 4,000 proteins, which makes it a daunting challenge to pick out proteins that might elicit a strong immune response if used as a vaccine.

In the new study, Bryson and his students set out to narrow the field of candidates by identifying TB proteins presented on the surface of infected human cells. When an immune cell such as a phagocyte is infected with Mycobacterium tuberculosis, some of the bacterial proteins get chopped into fragments called peptides, which are then displayed on the surface of the cell by MHC proteins. These MHC-peptide complexes act as a signal that can activate T cells.

MHCs, or major histocompatibility complexes, come in two types known as class I and class II. Class I MHCs activate killer T cells, while class II MHCs stimulate helper T cells. In human cells, there are three genes that can encode MHC-II proteins, and each of these comes in hundreds of variants. This means that any two people can have a very different repertoire of MHC-II molecules, which present different antigens.

“Instead of looking at all of those 4,000 TB proteins, we wanted to ask which of those proteins from TB actually end up being displayed to the rest of the immune system via MHC,” Bryson says. “If we could just answer that question, then we could design vaccines to match that.”

To try to answer the question, the researchers infected human phagocytes with Mycobacterium tuberculosis. After three days, they extracted MHC-peptide complexes from the cell surfaces, then identified the peptides using mass spectrometry.

Focusing on peptides bound to MHC-II, the researchers found 27 TB peptides, from 13 proteins, that appeared most often in the infected cells. Then, they further tested those peptides by exposing them to T cells donated by people who had previously been infected with TB.

They found that 24 of these peptides did elicit a T cell response in at least some of the samples. None of the proteins from which these peptides came worked for every single donor, but Bryson believes that a vaccine using a combination of these peptides would likely work for most people.

“In a perfect world, if you were trying to design a vaccine, you would pick one protein and that protein would be presented across every donor. It should work for every person,” Bryson says. “However, using our measurements, we’ve not yet found a TB protein that covers every donor we’ve analyzed thus far.”

Enter mRNA vaccines

Among the vaccine candidates that the researchers identified are several peptides from a class of proteins called type 7 secretion systems (T7SSs). Some of these peptides also turned up in an earlier study from Bryson’s lab on MHC-1.

“Type 7 secretion system substrates are a very small sliver of the overall TB proteome, but when you look at MHC class I or MHC class II, it seems as though the cells are preferentially presenting these,” Bryson says.

Two of the best-known of these proteins, EsxA and EsxB, are secreted by bacteria to help them escape from the membranes that phagocytes use to envelop them within the cell. Neither protein can break through the membrane on its own, but when joined together to form a heterodimer, they can poke holes, which also allow other T7SS proteins to escape.

To evaluate whether the proteins they identified could make a good vaccine, the researchers created mRNA vaccines encoding two protein sequences — EsxB and EsxG. The researchers designed several versions of the vaccine, which were targeted to different compartments within the cells.

The researchers then delivered this vaccine into human phagocytes, where they found that vaccines that targeted cell lysosomes — organelles that break down molecules — were the most effective. These vaccines induced 1,000 times more MHC presentation of TB peptides than any of the others.

They later found that the presentation was even higher if they added EsxA to the vaccine, because it allows the formation of the heterodimers that can poke through the lysosomal membrane.

The researchers currently have a mix of eight proteins that they believe could offer protection against TB for most people, but they are continuing to test the combination with blood samples from people around the world. They also hope to run additional studies to explore how much protection this vaccine offers in animal models. Tests in humans are likely several years away.

The research was funded by the MIT Center for Precision Cancer Research at the Koch Institute, the National Institutes of Health, the National Institute of Environmental Health Sciences, and the Frederick National Laboratory for Cancer Research.

A robot searching for workers trapped in a partially collapsed mine shaft must rapidly generate a map of the scene and identify its location within that scene as it navigates the treacherous terrain.

Researchers have recently started building powerful machine-learning models to perform this complex task using only images from the robot’s onboard cameras, but even the best models can only process a few images at a time. In a real-world disaster where every second counts, a search-and-rescue robot would need to quickly traverse large areas and process thousands of images to complete its mission.

To overcome this problem, MIT researchers drew on ideas from both recent artificial intelligence vision models and classical computer vision to develop a new system that can process an arbitrary number of images. Their system accurately generates 3D maps of complicated scenes like a crowded office corridor in a matter of seconds. 

The AI-driven system incrementally creates and aligns smaller submaps of the scene, which it stitches together to reconstruct a full 3D map while estimating the robot’s position in real-time.

Unlike many other approaches, their technique does not require calibrated cameras or an expert to tune a complex system implementation. The simpler nature of their approach, coupled with the speed and quality of the 3D reconstructions, would make it easier to scale up for real-world applications.

Beyond helping search-and-rescue robots navigate, this method could be used to make extended reality applications for wearable devices like VR headsets or enable industrial robots to quickly find and move goods inside a warehouse.

“For robots to accomplish increasingly complex tasks, they need much more complex map representations of the world around them. But at the same time, we don’t want to make it harder to implement these maps in practice. We’ve shown that it is possible to generate an accurate 3D reconstruction in a matter of seconds with a tool that works out of the box,” says Dominic Maggio, an MIT graduate student and lead author of a paper on this method.

Maggio is joined on the paper by postdoc Hyungtae Lim and senior author Luca Carlone, associate professor in MIT’s Department of Aeronautics and Astronautics (AeroAstro), principal investigator in the Laboratory for Information and Decision Systems (LIDS), and director of the MIT SPARK Laboratory. The research will be presented at the Conference on Neural Information Processing Systems.

Mapping out a solution

For years, researchers have been grappling with an essential element of robotic navigation called simultaneous localization and mapping (SLAM). In SLAM, a robot recreates a map of its environment while orienting itself within the space.

Traditional optimization methods for this task tend to fail in challenging scenes, or they require the robot’s onboard cameras to be calibrated beforehand. To avoid these pitfalls, researchers train machine-learning models to learn this task from data.

While they are simpler to implement, even the best models can only process about 60 camera images at a time, making them infeasible for applications where a robot needs to move quickly through a varied environment while processing thousands of images.

To solve this problem, the MIT researchers designed a system that generates smaller submaps of the scene instead of the entire map. Their method “glues” these submaps together into one overall 3D reconstruction. The model is still only processing a few images at a time, but the system can recreate larger scenes much faster by stitching smaller submaps together.

“This seemed like a very simple solution, but when I first tried it, I was surprised that it didn’t work that well,” Maggio says.

Searching for an explanation, he dug into computer vision research papers from the 1980s and 1990s. Through this analysis, Maggio realized that errors in the way the machine-learning models process images made aligning submaps a more complex problem.

Traditional methods align submaps by applying rotations and translations until they line up. But these new models can introduce some ambiguity into the submaps, which makes them harder to align. For instance, a 3D submap of a one side of a room might have walls that are slightly bent or stretched. Simply rotating and translating these deformed submaps to align them doesn’t work.

“We need to make sure all the submaps are deformed in a consistent way so we can align them well with each other,” Carlone explains.

A more flexible approach

Borrowing ideas from classical computer vision, the researchers developed a more flexible, mathematical technique that can represent all the deformations in these submaps. By applying mathematical transformations to each submap, this more flexible method can align them in a way that addresses the ambiguity.

Based on input images, the system outputs a 3D reconstruction of the scene and estimates of the camera locations, which the robot would use to localize itself in the space.

“Once Dominic had the intuition to bridge these two worlds — learning-based approaches and traditional optimization methods — the implementation was fairly straightforward,” Carlone says. “Coming up with something this effective and simple has potential for a lot of applications.

Their system performed faster with less reconstruction error than other methods, without requiring special cameras or additional tools to process data. The researchers generated close-to-real-time 3D reconstructions of complex scenes like the inside of the MIT Chapel using only short videos captured on a cell phone.

The average error in these 3D reconstructions was less than 5 centimeters.

In the future, the researchers want to make their method more reliable for especially complicated scenes and work toward implementing it on real robots in challenging settings.

“Knowing about traditional geometry pays off. If you understand deeply what is going on in the model, you can get much better results and make things much more scalable,” Carlone says.

This work is supported, in part, by the U.S. National Science Foundation, U.S. Office of Naval Research, and the National Research Foundation of Korea. Carlone, currently on sabbatical as an Amazon Scholar, completed this work before he joined Amazon.

What if clinicians could place tiny electronic chips in the brain that electrically stimulate a precise target, through a simple injection in the arm? This may someday help treat deadly or debilitating brain diseases, while eliminating surgery-related risks and costs.

MIT researchers have taken a major step toward making this scenario a reality. They developed microscopic, wireless bioelectronics that could travel through the body’s circulatory system and autonomously self-implant in a target region of the brain, where they would provide focused treatment.

In a study on mice, the researchers show that after injection, these miniscule implants can identify and travel to a specific brain region without the need for human guidance. Once there, they can be wirelessly powered to provide electrical stimulation to the precise area. Such stimulation, known as neuromodulation, has shown promise as a way to treat brain tumors and diseases like Alzheimer’s and multiple sclerosis.

Moreover, because the electronic devices are integrated with living, biological cells before being injected, they are not attacked by the body’s immune system and can cross the blood-brain barrier while leaving it intact. This maintains the barrier’s crucial protection of the brain.

The researchers demonstrated the use of this technology, which they call “circulatronics,” to target brain inflammation, a major factor in the progression of many neurological diseases. They show that the implants can provide localized neuromodulation deep inside the brain achieving high precision, to within several microns around the target area.

In addition, the biocompatible implants do not damage surrounding neurons.

While brain implants usually require hundreds of thousands of dollars in medical costs and risky surgical procedures, circulatronics technology holds the potential to make therapeutic brain implants accessible to all by eliminating the need for surgery, says Deblina Sarkar, the AT&T Career Development Associate Professor in the MIT Media Lab and MIT Center for Neurobiological Engineering, head of the Nano-Cybernetic Biotrek Lab, and senior author of a study on the work.

She is joined on the paper by lead author Shubham Yadav, an MIT graduate student; as well as others at MIT, Wellesley College, and Harvard University. The research appears today in Nature Biotechnology.

Hybrid implants

The team has been working on circulatronics for more than six years. The electronic devices, each about one-billionth the length of a grain of rice, are composed of organic semiconducting polymer layers sandwiched between metallic layers to create an electronic heterostructure.

They are fabricated using CMOS-compatible processes in the MIT.nano facilities, and then integrated with living cells to create cell-electronics hybrids. To do this, the researchers lift the devices off the silicon wafer on which they are fabricated, so they are free-floating in a solution.

“The electronics worked perfectly when they were attached to the substrate, but when we originally lifted them off, they didn’t work anymore. Solving that challenge took us more than a year,” Sarkar says.

Key to their operation is the high wireless power conversion efficiency of the tiny electronics. This enables the devices to work deep inside the brain and still harness enough energy for neuromodulation.

The researchers use a chemical reaction to bond the electronic devices to cells. In the new study, they fused the electronics with a type of immune cell called monocytes, which target areas of inflammation in the body. They also applied a fluorescent dye, allowing them to trace the devices as they crossed the intact blood-brain barrier and self-implanted in the target brain region.

While they explored brain inflammation in this study, the researchers hope to use different cell types and engineer the cells to target specific regions of the brain.

“Our cell-electronics hybrid fuses the versatility of electronics with the biological transport and biochemical sensing prowess of living cells,” Sarkar says. “The living cells camouflage the electronics so that they aren’t attacked by the body’s immune system and they can travel seamlessly through the bloodstream. This also enables them to squeeze through the intact blood-brain barrier without the need to invasively open it.”

Over the course of about four years, the team tried many methods to autonomously and noninvasively cross the blood-brain barrier before they perfected this cellular integration technique.

In addition, because the circulatronics devices are so tiny, they offer much higher precision than conventional electrodes. They can self-implant, leading to millions of microscopic stimulation sites that take the exact shape of the target region.

Their small size also enables the biocompatible devices to live alongside neurons without causing harmful effects. Through a series of biocompatibility tests, the researchers found that circulatronics can safely integrate among neurons without impacting the brain processes behind cognition or motion.

After the devices have self-implanted in the target region, a clinician or researcher uses an external transmitter to provide electromagnetic waves, in the form of near-infrared light, that power the technology and enable electrical stimulation of the neurons.

Targeting deadly diseases

The Sarkar lab is currently working on developing their technology to treat multiple diseases including brain cancer, Alzheimer’s disease, and chronic pain.

The tiny size and self-implantation capabilities of circulatronics devices could make them well-suited to treat brain cancers such as glioblastoma that cause tumors at multiple locations, some of which may be too small to identify with imaging techniques. They may also provide new avenues for treating especially deadly cancers like diffuse intrinsic pontine glioma, an aggressive type of tumor found in the brain stem that usually cannot be surgically removed.

“This is a platform technology and may be employed to treat multiple brain diseases and mental illnesses,” Sarkar says. “Also, this technology is not just confined to the brain but could also be extended to other parts of the body in future.”

The researchers hope to move the technology into clinical trials within three years through the recently launched startup Cahira Technologies.

They are also exploring integration of additional nanoelectronic circuits into their devices to enable functionalities including sensing, feedback based on-chip data analysis, and capabilities such as creating synthetic electronic neurons.

“Our tiny electronic devices seamlessly integrate with the neurons and co-live and co-exist with the brain cells creating a unique brain-computer symbiosis. We are working dedicatedly to employ this technology for treating neural diseases, where drugs or standard therapies fail, for alleviating human suffering and envision a future where humans could transcend beyond diseases and biological limitations,” says Sarkar.

One of the highest-risk components of nuclear waste is iodine-129 (I-129), which stays radioactive for millions of years and accumulates in human thyroids when ingested. In the U.S., nuclear waste containing I-129 is scheduled to be disposed of in deep underground repositories, which scientists say will sufficiently isolate it.

Meanwhile, across the globe, France routinely releases low-level radioactive effluents containing iodine-129 and other radionuclides into the ocean. France recycles its spent nuclear fuel, and the reprocessing plant discharges about 153 kilograms of iodine-129 each year, under the French regulatory limit.

Is dilution a good solution? What’s the best way to handle spent nuclear fuel? A new study by MIT researchers and their collaborators at national laboratories quantifies I-129 release under three different scenarios: the U.S. approach of disposing spent fuel directly in deep underground repositories, the French approach of dilution and release, and an approach that uses filters to capture I-129 and disposes of them in shallow underground waste repositories.

The researchers found France’s current practice of reprocessing releases about 90 percent of the waste’s I-129 into the biosphere. They found low levels of I-129 in ocean water around France and the U.K.’s former reprocessing sites, including the English Channel and North Sea. Although the low level of I-129 in the water in Europe is not considered to pose health risks, the U.S. approach of deep underground disposal leads to far less I-129 being released, the researchers found.

The researchers also investigated the effect of environmental regulations and technologies related to I-129 management, to illuminate the tradeoffs associated with different approaches around the world.

“Putting these pieces together to provide a comprehensive view of Iodine-129 is important,” says MIT Assistant Professor Haruko Wainwright, a first author on the paper who holds a joint appointment in the departments of Nuclear Science and Engineering and of Civil and Environmental Engineering. “There are scientists that spend their lives trying to clean up iodine-129 at contaminated sites. These scientists are sometimes shocked to learn some countries are releasing so much iodine-129. This work also provides a life-cycle perspective. We’re not just looking at final disposal and solid waste, but also when and where release is happening. It puts all the pieces together.”

MIT graduate student Kate Whiteaker SM ’24 led many of the analyses with Wainwright. Their co-authors are Hansell Gonzalez-Raymat, Miles Denham, Ian Pegg, Daniel Kaplan, Nikolla Qafoku, David Wilson, Shelly Wilson, and Carol Eddy-Dilek. The study appears today in Nature Sustainability.

Managing waste

Iodine-129 is often a key focus for scientists and engineers as they conduct safety assessments of nuclear waste disposal sites around the world. It has a half-life of 15.7 million years, high environmental mobility, and could potentially cause cancers if ingested. The U.S. sets a strict limit on how much I-129 can be released and how much I-129 can be in drinking water — 5.66 nanograms per liter, the lowest such level of any radionuclides.

“Iodine-129 is very mobile, so it is usually the highest-dose contributor in safety assessments,” Wainwright says.

For the study, the researchers calculated the release of I-129 across three different waste management strategies by combining data from current and former reprocessing sites as well as repository assessment models and simulations.

The authors defined the environmental impact as the release of I-129 into the biosphere that humans could be exposed to, as well as its concentrations in surface water. They measured I-129 release per the total electrical energy generated by a 1-gigawatt power plant over one year, denoted as kg/GWe.y.

Under the U.S. approach of deep underground disposal with barrier systems, assuming the barrier canisters fail at 1,000 years (a conservative estimate), the researchers found 2.14 x 10–8 kg/GWe.y of I-129 would be released between 1,000 and 1 million years from today.

They estimate that 4.51 kg/GWe.y of I-129, or 91 percent of the total, would be released into the biosphere in the scenario where fuel is reprocessed and the effluents are diluted and released. About 3.3 percent of I-129 is captured by gas filters, which are then disposed of in shallow subsurfaces as low-level radioactive waste. A further 5.2 percent remains in the waste stream of the reprocessing plant, which is then disposed of as high-level radioactive waste.

If the waste is recycled with gas filters to directly capture I-129, 0.05 kg/GWe.y of the I-129 is released, while 94 percent is disposed of in the low-level disposal sites. For shallow disposal, some kind of human disruption and intrusion is assumed to occur after government or institutional control expires (typically 100-1,000 years). That results in a potential release of the disposed amount to the environment after the control period.

Overall, the current practice of recycling spent nuclear fuel releases the majority of I-129 into the environment today, while the direct disposal of spent fuel releases around 1/100,000,000 that amount over 1 million years. When the gas filters are used to capture I-129, the majority of I-129 goes to shallow underground repositories, which could be accidentally released through human intrusion down the line.

The researchers also quantified the concentration of I-129 in different surface waters near current and former fuel reprocessing facilities, including the English Channel and the North Sea near reprocessing plants in France and U.K. They also analyzed the U.S. Columbia River downstream of a site in Washington state where material for nuclear weapons was produced during the Cold War, and they studied a similar site in South Carolina. The researchers found far higher concentrations of I-129 within the South Carolina site, where the low-level radioactive effluents were released far from major rivers and hence resulted in less dilution in the environment.

“We wanted to quantify the environmental factors and the impact of dilution, which in this case affected concentrations more than discharge amounts,” Wainwright says. “Someone might take our results to say dilution still works: It’s reducing the contaminant concentration and spreading it over a large area. On the other hand, in the U.S., imperfect disposal has led to locally higher surface water concentrations. This provides a cautionary tale that disposal could concentrate contaminants, and should be carefully designed to protect local communities.”

Fuel cycles and policy

Wainwright doesn’t want her findings to dissuade countries from recycling nuclear fuel. She says countries like Japan plan to use increased filtration to capture I-129 when they reprocess spent fuel. Filters with I-129 can be disposed of as low-level waste under U.S. regulations.

“Since I-129 is an internal carcinogen without strong penetrating radiation, shallow underground disposal would be appropriate in line with other hazardous waste,” Wainwright says. “The history of environmental protection since the 1960s is shifting from waste dumping and release to isolation. But there are still industries that release waste into the air and water. We have seen that they often end up causing issues in our daily life — such as CO2, mercury, PFAS and others — especially when there are many sources or when bioaccumulation happens. The nuclear community has been leading in waste isolation strategies and technologies since the 1950s. These efforts should be further enhanced and accelerated. But at the same time, if someone does not choose nuclear energy because of waste issues, it would encourage other industries with much lower environmental standards.”

The work was supported by MIT’s Climate Fast Forward Faculty Fund and the U.S. Department of Energy.

Although heart cells and skin cells contain identical instructions for creating proteins encoded in their DNA, they’re able to fill such disparate niches because molecular machinery can cut out and stitch together different segments of those instructions to create endlessly unique combinations.

The ingenuity of using the same genes in different ways is made possible by a process called splicing and is controlled by splicing factors; which splicing factors a cell employs determines what sets of instructions that cell produces, which, in turn, gives rise to proteins that allow cells to fulfill different functions. 

In an open-access paper published today in Nature Biotechnology, researchers in the MIT Department of Biology outlined a framework for parsing the complex relationship between sequences and splicing regulation to investigate the regulatory activities of splicing factors, creating models that can be applied to interpret and predict splicing regulation across different cell types, and even different species. Called Knockdown Activity and Target Models from Additive regression Predictions, KATMAP draws on experimental data from disrupting the expression of a splicing factor and information on which sequences the splicing factor interacts with to predict its likely targets. 

Aside from the benefits of a better understanding of gene regulation, splicing mutations — either in the gene that is spliced or in the splicing factor itself — can give rise to diseases such as cancer by altering how genes are expressed, leading to the creation or accumulation of faulty or mutated proteins. This information is critical for developing therapeutic treatments for those diseases. The researchers also demonstrated that KATMAP can potentially be used to predict whether synthetic nucleic acids, a promising treatment option for disorders including a subset of muscular atrophy and epilepsy disorders, affect splicing.

Perturbing splicing 

In eukaryotic cells, including our own, splicing occurs after DNA is transcribed to produce an RNA copy of a gene, which contains both coding and non-coding regions of RNA. The noncoding intron regions are removed, and the coding exon segments are spliced back together to make a near-final blueprint, which can then be translated into a protein. 

According to first author Michael P. McGurk, a postdoc in the lab of MIT Professor Christopher Burge, previous approaches could provide an average picture of regulation, but could not necessarily predict the regulation of splicing factors at particular exons in particular genes.

KATMAP draws on RNA sequencing data generated from perturbation experiments, which alter the expression level of a regulatory factor by either overexpressing it or knocking down its levels. The consequences of overexpression or knockdown are that the genes regulated by the splicing factor should exhibit different levels of splicing after perturbation, which helps the model identify the splicing factor’s targets. 

Cells, however, are complex, interconnected systems, where one small change can cause a cascade of effects. KATMAP is also able to distinguish between direct targets from indirect, downstream impacts by incorporating known information about the sequence the splicing factor is likely to interact with, referred to as a binding site or binding motif.

“In our analyses, we identify predicted targets as exons that have binding sites for this particular factor in the regions where this model thinks they need to be to impact regulation,” McGurk says, while non-targets may be affected by perturbation but don’t have the likely appropriate binding sites nearby. 

This is especially helpful for splicing factors that aren’t as well-studied. 

“One of our goals with KATMAP was to try to make the model general enough that it can learn what it needs to assume for particular factors, like how similar the binding site has to be to the known motif or how regulatory activity changes with the distance of the binding sites from the splice sites,” McGurk says. 

Starting simple

Although predictive models can be very powerful at presenting possible hypotheses, many are considered “black boxes,” meaning the rationale that gives rise to their conclusions is unclear. KATMAP, on the other hand, is an interpretable model that enables researchers to quickly generate hypotheses and interpret splicing patterns in terms of regulatory factors while also understanding how the predictions were made. 

“I don’t just want to predict things, I want to explain and understand,” McGurk says. “We set up the model to learn from existing information about splicing and binding, which gives us biologically interpretable parameters.” 

The researchers did have to make some simplifying assumptions in order to develop the model. KATMAP considers only one splicing factor at a time, although it is possible for splicing factors to work in concert with one another. The RNA target sequence could also be folded in such a way that the factor wouldn’t be able to access a predicted binding site, so the site is present but not utilized.

“When you try to build up complete pictures of complex phenomena, it’s usually best to start simple,” McGurk says. “A model that only considers one splicing factor at a time is a good starting point.” 

David McWaters, another postdoc in the Burge Lab and a co-author on the paper, conducted key experiments to test and validate that aspect of the KATMAP model.

Future directions

The Burge lab is collaborating with researchers at Dana-Farber Cancer Institute to apply KATMAP to the question of how splicing factors are altered in disease contexts, as well as with other researchers at MIT as part of an MIT HEALS grant to model splicing factor changes in stress responses. McGurk also hopes to extend the model to incorporate cooperative regulation for splicing factors that work together. 

“We’re still in a very exploratory phase, but I would like to be able to apply these models to try to understand splicing regulation in disease or development. In terms of variation of splicing factors, they are related, and we need to understand both,” McGurk says.

Burge, the Uncas (1923) and Helen Whitaker Professor and senior author of the paper, will continue to work on generalizing this approach to build interpretable models for other aspects of gene regulation.

“We now have a tool that can learn the pattern of activity of a splicing factor from types of data that can be readily generated for any factor of interest,” says Burge, who is also an extra-mural member of the Koch Institute for Integrative Cancer Research and an associate member of the Broad Institute of MIT and Harvard. “As we build up more of these models, we’ll be better able to infer which splicing factors have altered activity in a disease state from transcriptomic data, to help understand which splicing factors are driving pathology.”

MIT engineers have developed a flexible drug-delivery patch that can be placed on the heart after a heart attack to help promote healing and regeneration of cardiac tissue.

The new patch is designed to carry several different drugs that can be released at different times, on a pre-programmed schedule. In a study of rats, the researchers showed that this treatment reduced the amount of damaged heart tissue by 50 percent and significantly improved cardiac function.

If approved for use in humans, this type of patch could help heart attack victims recover more of their cardiac function than is now possible, the researchers say.

“When someone suffers a major heart attack, the damaged cardiac tissue doesn’t regenerate effectively, leading to a permanent loss of heart function. The tissue that was damaged doesn’t recover,” says Ana Jaklenec, a principal investigator at MIT’s Koch Institute for Integrative Cancer Research. “Our goal is to restore that function and help people regain a stronger, more resilient heart after a myocardial infarction.”

Jaklenec and Robert Langer, the David H. Koch Institute Professor at MIT and a member of the Koch Institute, are the senior authors of the new study, which appears today in Cell Biomaterials. Former MIT postdoc Erika Wangis the lead author of the paper.

Programmed drug delivery

After a heart attack, many patients end up having bypass surgery, which improves blood flow to the heart but doesn’t repair the cardiac tissue that was damaged. In the new study, the MIT team wanted to create a patch that could be applied to the heart at the same time that the surgery is performed.

This patch, they hoped, could deliver drugs over an extended time period to promote tissue healing. Many diseases, including heart conditions, require phase-specific treatment, but most systems release drugs all at once. Timed delivery better synchronizes therapy with recovery.

“We wanted to see if it’s possible to deliver a precisely orchestrated therapeutic intervention to help heal the heart, right at the site of damage, while the surgeon is already performing open-heart surgery,” Jaklenec says.

To achieve this, the researchers set out to adapt drug-delivery microparticles they had previously developed, which consist of capsules similar to tiny coffee cups with lids. These capsules are made from a polymer called PLGA and can be sealed with a drug inside.

By changing the molecular weight of the polymers used to form the lids, the researchers can control how quickly they degrade, which enables them to program the particles to release their contents at specific times. For this application, the researchers designed particles that break down during days 1-3, days 7-9, and days 12-14 after implantation.

This allowed them to devise a regimen of three drugs that promote heart healing in different ways. The first set of particles release neuregulin-1, a growth factor that helps to prevent cell death. At the next time point, particles release VEGF, a growth factor that promotes formation of blood vessels surrounding the heart. The last batch of particles releases a small molecule drug called GW788388, which inhibits the formation of scar tissue that can occur following a heart attack.

“When tissue regenerates, it follows a carefully timed series of steps,” Jaklenec says. “Dr. Wang created a system that delivers key components at just the right time, in the sequence that the body naturally uses to heal.”

The researchers embedded rows of these particles into thin sheets of a tough but flexible hydrogel, similar to a contact lens. This hydrogel is made from alginate and PEGDA, two biocompatible polymers that eventually break down in the body. For this study, the researchers created compact, miniature patches only a few millimeters across.

“We encapsulate arrays of these particles in a hydrogel patch, and then we can surgically implant this patch into the heart. In this way, we’re really programming the treatment into this material,” Wang says.

Better heart function

Once they created these patches, the researchers tested them on spheres of heart tissue that included cardiomyocytes generated from induced pluripotent stem cells. These spheres also included endothelial cells and human ventricular cardiac fibroblasts, which are also important components of the heart.

The researchers exposed those spheres to low-oxygen conditions, mimicking the effects of a heart attack, then placed the patches over them. They found that the patches promoted blood vessel growth, helped more cells to survive, and reduced the amount of fibrosis that developed.

In tests in a rat model of heart attack, the researchers also saw significant improvements following treatment with the patch. Compared to no treatment or IV injection of the same drugs, animals treated with the patch showed 33 percent higher survival rates, a 50 percent reduction in the amount of damaged tissue, and significantly increased cardiac output.

The researchers showed that the patches would eventually dissolve over time, becoming a very thin layer over the course of a year without disrupting the heart’s mechanical function.

“This is an important way to combine drug delivery and biomaterials to potentially new treatments for patients,” Langer says.

Of the drugs tested in this study, neuregulin-1 and VEGF have been tested in clinical trials to treat heart conditions, but GW788388 has only been explored in animal models. The researchers now hope to test their patches in additional animal models in hopes of running a clinical trial in the future.

The current version of the patch needs to be implanted surgically, but the researchers are exploring the possibility of incorporating these microparticles into stents that could be inserted into arteries to deliver drugs on a programmed schedule.

Other authors of the paper include Elizabeth Calle, Binbin Ying, Behnaz Eshaghi, Linzixuan Zhang, Xin Yang, Stacey Qiaohui Lin, Jooli Han, Alanna Backx, Yuting Huang, Sevinj Mursalova, Chuhan Joyce Qi, and Yi Liu.

The researchers were supported by the Natural Sciences and Engineering Research Council of Canada and the U.S. National Heart, Lung, and Blood Institute.

More than 70 aircraft are struck by lightning every day. If you happen to be flying when a strike occurs, chances are you won’t feel a thing, thanks to lightning protection measures that are embedded in key zones throughout the aircraft.

Lightning protection systems work well, largely because they are designed for planes with a “tube-and-wing” structure, a simple geometry common to most aircraft today. But future airplanes may not look and fly the same way. The aviation industry is exploring new designs, including blended-wing bodies and truss-braced wings, partly to reduce fuel and weight costs. But researchers don’t yet know how these unconventional designs might respond to lightning strikes.

MIT aerospace engineers are hoping to change that with a new physics-based approach that predicts how lightning would sweep across a plane with any design. The tool then generates a zoning map highlighting sections of an aircraft that would require various degrees of lightning protection, given how they are likely to experience a strike.

“People are starting to conceive aircraft that look very different from what we’re used to, and we can’t apply exactly what we know from historical data to these new configurations because they’re just too different,” says Carmen Guerra-Garcia, associate professor of aeronautics and astronautics (AeroAstro) at MIT. “Physics-based methods are universal. They’re agnostic to the type of geometry or vehicle. This is the path forward to be able to do this lightning zoning and protect future aircraft.”

She and her colleagues report their results in a study appearing this week in IEEE Access. The study’s first author is AeroAstro graduate student Nathanael Jenkins. Other co-authors include Louisa Michael and Benjamin Westin of Boeing Research and Technology.

First strike

When lightning strikes, it first attaches to a part of a plane — typically a sharp edge or extremity — and hangs on for up to a second. During this brief flash, the plane continues speeding through the air, causing the lightning current to “sweep” over parts of its surface, potentially changing in intensity and re-attaching at certain points where the intense current flow could damage vulnerable sections of an aircraft.

In previous work, Guerra-Garcia’s group developed a model to predict the parts of a plane where lightning is most likely to first connect. That work, led by graduate student Sam Austin, established a starting point for the team’s new work, which aims to predict how and where the lightning will then sweep over the plane’s surface. The team next converted their lightning sweep predictions into zoning maps to identify vulnerable regions requiring certain levels of protection.

A typical tube-and-wing plane is divided into three main zones, as classified by the aviation industry. Each zone has a clear description of the level of current it must withstand in order to be certified for flight. Parts of a plane that are more likely to be hit by lightning are generally classified as zone 1 and require more protection, which can include embedded metal foil in the skin of the airplane that conducts away a lightning current.

To date, an airplane’s lightning zones have been determined over many years of flight inspections after lightning strikes and fine-tuning of protection measures. Guerra-Garcia and her colleagues looked to develop a zoning approach based on physics, rather than historical flight data. Such a physics-based mapping could be applied to any shape of aircraft, such as unconventional and largely untested designs, to identify regions that really require reinforcement.

“Protecting aircraft from lightning is heavy,” Jenkins says. “Embedding copper mesh or foil throughout an aircraft is an added weight penalty. And if we had the greatest level of protection for every part of the plane’s surface, the plane would weigh far too much. So zoning is about trying to optimize the weight of the system while also having it be as safe as possible.”

In the zone

For their new approach, the team developed a model to predict the pattern of lightning sweep and the corresponding lightning protection zones, for a given airplane geometry. Starting with a specific airplane shape — in their case, a typical tube-and-wing structure — the researchers simulated the fluid dynamics, or how air would flow around a plane, given a certain speed, altitude, and pitch angle. They also incorporated their previous model that predicts the places where lightning is more likely to initially attach.

For each initial attachment point, the team simulated tens of thousands of potential lightning arcs, or angles from which the current strikes the plane. They then ran the model forward to predict how the tens of thousands of potential strikes would follow the air flow across the plane’s surface. These runs produced a statistical representation of where lightning, striking a specific point on a plane, is likely to flow and potentially cause damage. The team converted this statistical representation into a map of zones of varying vulnerability.

They validated the method on a conventional tube-and-wing structure, showing that the zoning maps generated by the physics-based approach were consistent with what the aviation industry has determined over decades of fine-tuning.

“We now have a physics-based tool that provides some metrics like the probability of lightning attachment and dwell time, which is how long an arc will linger at a specific point,” Guerra-Garcia explains. “We convert those physics metrics into zoning maps to show, if I’m in this red region, the lightning arc will stay for a long time, so that region needs to be heavily protected.”

The team is starting to apply the approach to new geometries, such as blended-wing designs and truss-braced structures. The researchers envision that the tool can help designers incorporate safe and efficient lightning-protection systems early on in the design process.

“Lightning is incredible and terrifying at the same time, and I have full confidence in flying on planes at the moment,” Jenkins says. “I want to have that same confidence in 20 years’ time. So, we need a new way to zone aircraft.”

“With physics-based methods like the ones developed with professor Guerra-Garcia’s group we have the opportunity to shape industry standards and as an industry rely on the underlying physics to develop guidelines for aircraft certification through simulation,” says co-author Louisa Michael of Boeing Technology Innovation. Currently, we are engaging with industrial committees to propose these methods to be included in Aerospace Recommended Practices.”

“Zoning unconventional aircraft is not an easy task,” adds co-author Ben Westin of Boeing Technology Innovation. “But these methods will allow us to confidently identify which threat levels each part of the aircraft needs to be protected against and certified for, and they give our design engineers a platform to do their best work to optimize aircraft design.”

Beyond airplanes, Guerra-Garcia is looking at ways to adapt the lightning protection model to other technologies, including wind turbines.

“About 60 percent of blade losses are due to lightning and will become worse as we move offshore because wind turbines will be even bigger and more susceptible to upward lightning,” she says. “They have many of the same challenges of a flowing gas environment. It’s more complex, and we will apply this same sort of methodology to this space.”

This research was funded, in part, by the Boeing Company.

Quantum computers have the potential to model new molecules and weather patterns better than any computer today. They may also one day accelerate artificial intelligence algorithms at a much lower energy footprint. But anyone interested in using quantum computers faces a steep learning curve that starts with getting access to quantum devices and then figuring out one of the many quantum software programs on the market.

Now qBraid, founded by Kanav Setia and Jason Necaise ’20, is providing a gateway to quantum computing with a platform that gives users access to the leading quantum devices and software. Users can log on to qBraid’s cloud-based interface and connect with quantum devices and other computing resources from leading companies like Nvidia, Microsoft, and IBM. In a few clicks, they can start coding or deploy cutting-edge software that works across devices.

“The mission is to take you from not knowing anything about quantum computing to running your first program on these amazing machines in less than 10 minutes,” Setia says. “We’re a one-stop platform that gives access to everything the quantum ecosystem has to offer. Our goal is to enable anyone — whether they’re enterprise customers, academics, or individual users — to build and ultimately deploy applications.”

Since its founding in June of 2020, qBraid has helped more than 20,000 people in more than 120 countries deploy code on quantum devices. That traction is ultimately helping to drive innovation in a nascent industry that’s expected to play a key role in our future.

“This lowers the barrier to entry for a lot of newcomers,” Setia says. “They can be up and running in a few minutes instead of a few weeks. That’s why we’ve gotten so much adoption around the world. We’re one of the most popular platforms for accessing quantum software and hardware.”

A quantum “software sandbox”

Setia met Necaise while the two interned at IBM. At the time, Necaise was an undergraduate at MIT majoring in physics, while Setia was at Dartmouth College. The two enjoyed working together, and Necaise said if Setia ever started a company, he’d be interested in joining.

A few months later, Setia decided to take him up on the offer. At Dartmouth, Setia had taken one of the first applied quantum computing classes, but students spent weeks struggling to install all the necessary software programs before they could even start coding.

“We hadn’t even gotten close to developing any useful algorithms,” Seita said. “The idea for qBraid was, ‘Why don’t we build a software sandbox in the cloud and give people an easy programming setup out of the box?’ Connection with the hardware would already be done.”

The founders received early support from the MIT Sandbox Innovation Fund and took part in the delta v summer startup accelerator run by the Martin Trust Center for MIT Entrepreneurship.

“Both programs provided us with very strong mentorship,” Setia says. “They give you frameworks on what a startup should look like, and they bring in some of the smartest people in the world to mentor you — people you’d never have access to otherwise.”

Necaise left the company in 2021. Setia, meanwhile, continued to find problems with quantum software outside of the classroom.

“This is a massive bottleneck,” Setia says. “I’d worked on several quantum software programs that pushed out updates or changes, and suddenly all hell broke loose on my codebase. I’d spend two to four weeks jostling with these updates that had almost nothing to do with the quantum algorithms I was working on.”

QBraid started as a platform with pre-installed software that let developers start writing code immediately. The company also added support for version-controlled quantum software so developers could build applications on top without worrying about changes. Over time, qBraid added connections to quantum computers and tools that lets quantum programs run across different devices.

“The pitch was you don’t need to manage a bunch of software or a whole bunch of cloud accounts,” Setia says. “We’re a single platform: the quantum cloud.”

QBraid also launched qBook, a learning platform that offers interactive courses in quantum computing.

“If you see a piece of code you like, you just click play and the code runs,” Setia says. “You can run a whole bunch of code, modify it on the fly, and you can understand how it works. It runs on laptops, iPads, and phones. A significant portion of our users are from developing countries, and they’re developing applications from their phones.”

Democratizing quantum computing

Today qBraid’s 20,000 users come from over 400 universities and 100 companies around the world. As qBraid’s user base has grown, the company went from integrating quantum computers onto their platform from the outside to creating a quantum operating system, qBraid-OS, that is currently being used by four leading quantum companies.

“We are productizing these quantum computers,” Setia explains. “Many quantum companies are realizing they want to focus their energy completely on the hardware, with us productizing their infrastructure. We’re like the operating system for quantum computers.”

People are using qBraid to build quantum applications in AI and machine learning, to discover new molecules or develop new drugs, and to develop applications in finance and cybersecurity. With every new use case, Setia says qBraid is democratizing quantum computing to create the quantum workforce that will continue to advance the field.

“[In 2018], an article in The New York Times said there were possibly less than 1,000 people in the world that could be called experts in quantum programming,” Setia says. “A lot of people want to access these cutting-edge machines, but they don’t have the right software backgrounds. They are just getting started and want to play with algorithms. QBraid gives those people an easy programming setup out of the box.”

With the rapid advancement of generative artificial intelligence, teachers and school leaders are looking for answers to complicated questions about successfully integrating technology into lessons, while also ensuring students actually learn what they’re trying to teach. 

Justin Reich, an associate professor in MIT’s Comparative Media Studies/Writing program, hopes a new guidebook published by the MIT Teaching Systems Lab can support K-12 educators as they determine what AI policies or guidelines to craft.

“Throughout my career, I’ve tried to be a person who researches education and technology and translates findings for people who work in the field,” says Reich. “When tricky things come along I try to jump in and be helpful.” 

“A Guide to AI in Schools: Perspectives for the Perplexed,” published this fall, was developed with the support of an expert advisory panel and other researchers. The project includes input from more than 100 students and teachers from around the United States, sharing their experiences teaching and learning with new generative AI tools. 

“We’re trying to advocate for an ethos of humility as we examine AI in schools,” Reich says. “We’re sharing some examples from educators about how they’re using AI in interesting ways, some of which might prove sturdy and some of which might prove faulty. And we won’t know which is which for a long time.”

Finding answers to AI and education questions

The guidebook attempts to help K-12 educators, students, school leaders, policymakers, and others collect and share information, experiences, and resources. AI’s arrival has left schools scrambling to respond to multiple challenges, like how to ensure academic integrity and maintain data privacy. 

Reich cautions that the guidebook is not meant to be prescriptive or definitive, but something that will help spark thought and discussion. 

“Writing a guidebook on generative AI in schools in 2025 is a little bit like writing a guidebook of aviation in 1905,” the guidebook’s authors note. “No one in 2025 can say how best to manage AI in schools.”

Schools are also struggling to measure how student learning loss looks in the age of AI. “How does bypassing productive thinking with AI look in practice?” Reich asks. “If we think teachers provide content and context to support learning and students no longer perform the exercises housing the content and providing the context, that’s a serious problem.”

Reich invites people directly impacted by AI to help develop solutions to the challenges its ubiquity presents. “It’s like observing a conversation in the teacher’s lounge and inviting students, parents, and other people to participate about how teachers think about AI,” he says, “what they are seeing in their classrooms, and what they’ve tried and how it went.”

The guidebook, in Reich’s view, is ultimately a collection of hypotheses expressed in interviews with teachers: well-informed, initial guesses about the paths that schools could follow in the years ahead. 

Producing educator resources in a podcast

In addition to the guidebook, the Teaching Systems Lab also recently produced “The Homework Machine,” a seven-part series from the Teachlab podcast that explores how AI is reshaping K-12 education. 

Reich produced the podcast in collaboration with journalist Jesse Dukes. Each episode tackles a specific area, asking important questions about challenges related to issues like AI adoption, poetry as a tool for student engagement, post-Covid learning loss, pedagogy, and book bans. The podcast allows Reich to share timely information about education-related updates and collaborate with people interested in helping further the work.

“The academic publishing cycle doesn’t lend itself to helping people with near-term challenges like those AI presents,” Reich says. “Peer review takes a long time, and the research produced isn’t always in a form that’s helpful to educators.” Schools and districts are grappling with AI in real time, bypassing time-tested quality control measures. 

The podcast can help reduce the time it takes to share, test, and evaluate AI-related solutions to new challenges, which could prove useful in creating training and resources.  

“We hope the podcast will spark thought and discussion, allowing people to draw from others’ experiences,” Reich says.

The podcast was also produced into an hour-long radio special, which was broadcast by public radio stations across the country.

“We’re fumbling around in the dark”

Reich is direct in his assessment of where we are with understanding AI and its impacts on education. “We’re fumbling around in the dark,” he says, recalling past attempts to quickly integrate new tech into classrooms. These failures, Reich suggests, highlight the importance of patience and humility as AI research continues. “AI bypassed normal procurement processes in education; it just showed up on kids’ phones,” he notes. 

“We’ve been really wrong about tech in the past,” Reich says. Despite districts’ spending on tools like smartboards, for example, research indicates there’s no evidence that they improve learning or outcomes. In a new article for article for The Conversation, he argues that early teacher guidance in areas like web literacy has produced bad advice that still exists in our educational system. “We taught students and educators not to trust Wikipedia,” he recalls, “and to search for website credibility markers, both of which turned out to be incorrect.” Reich wants to avoid a similar rush to judgment on AI, recommending that we avoid guessing at AI-enabled instructional strategies.

These challenges, coupled with potential and observed student impacts, significantly raise the stakes for schools and students’ families in the AI race. “Education technology always provokes teacher anxiety,” Reich notes, “but the breadth of AI-related concerns is much greater than in other tech-related areas.” 

The dawn of the AI age is different from how we’ve previously received tech into our classrooms, Reich says. AI wasn’t adopted like other tech. It simply arrived. It’s now upending educational models and, in some cases, complicating efforts to improve student outcomes.

Reich is quick to point out that there are no clear, definitive answers on effective AI implementation and use in classrooms; those answers don’t currently exist. Each of the resources Reich helped develop invite engagement from the audiences they target, aggregating valuable responses others might find useful.

“We can develop long-term solutions to schools’ AI challenges, but it will take time and work,” he says. “AI isn’t like learning to tie knots; we don’t know what AI is, or is going to be, yet.” 

Reich also recommends learning more about AI implementation from a variety of sources. “Decentralized pockets of learning can help us test ideas, search for themes, and collect evidence on what works,” he says. “We need to know if learning is actually better with AI.” 

While teachers don’t get to choose regarding AI’s existence, Reich believes it’s important that we solicit their input and involve students and other stakeholders to help develop solutions that improve learning and outcomes. 

“Let’s race to answers that are right, not first,” Reich says.

A recent study from Oregon State University estimated that more than 3,500 animal species are at risk of extinction because of factors including habitat alterations, natural resources being overexploited, and climate change.

To better understand these changes and protect vulnerable wildlife, conservationists like MIT PhD student and Computer Science and Artificial Intelligence Laboratory (CSAIL) researcher Justin Kay are developing computer vision algorithms that carefully monitor animal populations. A member of the lab of MIT Department of Electrical Engineering and Computer Science assistant professor and CSAIL principal investigator Sara Beery, Kay is currently working on tracking salmon in the Pacific Northwest, where they provide crucial nutrients to predators like birds and bears, while managing the population of prey, like bugs.

With all that wildlife data, though, researchers have lots of information to sort through and many AI models to choose from to analyze it all. Kay and his colleagues at CSAIL and the University of Massachusetts Amherst are developing AI methods that make this data-crunching process much more efficient, including a new approach called “consensus-driven active model selection” (or “CODA”) that helps conservationists choose which AI model to use. Their work was named a Highlight Paper at the International Conference on Computer Vision (ICCV) in October.

That research was supported, in part, by the National Science Foundation, Natural Sciences and Engineering Research Council of Canada, and Abdul Latif Jameel Water and Food Systems Lab (J-WAFS). Here, Kay discusses this project, among other conservation efforts.

Q: In your paper, you pose the question of which AI models will perform the best on a particular dataset. With as many as 1.9 million pre-trained models available in the HuggingFace Models repository alone, how does CODA help us address that challenge?

A: Until recently, using AI for data analysis has typically meant training your own model. This requires significant effort to collect and annotate a representative training dataset, as well as iteratively train and validate models. You also need a certain technical skill set to run and modify AI training code. The way people interact with AI is changing, though — in particular, there are now millions of publicly available pre-trained models that can perform a variety of predictive tasks very well. This potentially enables people to use AI to analyze their data without developing their own model, simply by downloading an existing model with the capabilities they need. But this poses a new challenge: Which model, of the millions available, should they use to analyze their data? 

Typically, answering this model selection question also requires you to spend a lot of time collecting and annotating a large dataset, albeit for testing models rather than training them. This is especially true for real applications where user needs are specific, data distributions are imbalanced and constantly changing, and model performance may be inconsistent across samples. Our goal with CODA was to substantially reduce this effort. We do this by making the data annotation process “active.” Instead of requiring users to bulk-annotate a large test dataset all at once, in active model selection we make the process interactive, guiding users to annotate the most informative data points in their raw data. This is remarkably effective, often requiring users to annotate as few as 25 examples to identify the best model from their set of candidates. 

We’re very excited about CODA offering a new perspective on how to best utilize human effort in the development and deployment of machine-learning (ML) systems. As AI models become more commonplace, our work emphasizes the value of focusing effort on robust evaluation pipelines, rather than solely on training.

Q: You applied the CODA method to classifying wildlife in images. Why did it perform so well, and what role can systems like this have in monitoring ecosystems in the future?

A: One key insight was that when considering a collection of candidate AI models, the consensus of all of their predictions is more informative than any individual model’s predictions. This can be seen as a sort of “wisdom of the crowd:” On average, pooling the votes of all models gives you a decent prior over what the labels of individual data points in your raw dataset should be. Our approach with CODA is based on estimating a “confusion matrix” for each AI model — given the true label for some data point is class X, what is the probability that an individual model predicts class X, Y, or Z? This creates informative dependencies between all of the candidate models, the categories you want to label, and the unlabeled points in your dataset.

Consider an example application where you are a wildlife ecologist who has just collected a dataset containing potentially hundreds of thousands of images from cameras deployed in the wild. You want to know what species are in these images, a time-consuming task that computer vision classifiers can help automate. You are trying to decide which species classification model to run on your data. If you have labeled 50 images of tigers so far, and some model has performed well on those 50 images, you can be pretty confident it will perform well on the remainder of the (currently unlabeled) images of tigers in your raw dataset as well. You also know that when that model predicts some image contains a tiger, it is likely to be correct, and therefore that any model that predicts a different label for that image is more likely to be wrong. You can use all these interdependencies to construct probabilistic estimates of each model’s confusion matrix, as well as a probability distribution over which model has the highest accuracy on the overall dataset. These design choices allow us to make more informed choices over which data points to label and ultimately are the reason why CODA performs model selection much more efficiently than past work.

There are also a lot of exciting possibilities for building on top of our work. We think there may be even better ways of constructing informative priors for model selection based on domain expertise — for instance, if it is already known that one model performs exceptionally well on some subset of classes or poorly on others. There are also opportunities to extend the framework to support more complex machine-learning tasks and more sophisticated probabilistic models of performance. We hope our work can provide inspiration and a starting point for other researchers to keep pushing the state of the art.

Q: You work in the Beerylab, led by Sara Beery, where researchers are combining the pattern-recognition capabilities of machine-learning algorithms with computer vision technology to monitor wildlife. What are some other ways your team is tracking and analyzing the natural world, beyond CODA?

A: The lab is a really exciting place to work, and new projects are emerging all the time. We have ongoing projects monitoring coral reefs with drones, re-identifying individual elephants over time, and fusing multi-modal Earth observation data from satellites and in-situ cameras, just to name a few. Broadly, we look at emerging technologies for biodiversity monitoring and try to understand where the data analysis bottlenecks are, and develop new computer vision and machine-learning approaches that address those problems in a widely applicable way. It’s an exciting way of approaching problems that sort of targets the “meta-questions” underlying particular data challenges we face. 

The computer vision algorithms I’ve worked on that count migrating salmon in underwater sonar video are examples of that work. We often deal with shifting data distributions, even as we try to construct the most diverse training datasets we can. We always encounter something new when we deploy a new camera, and this tends to degrade the performance of computer vision algorithms. This is one instance of a general problem in machine learning called domain adaptation, but when we tried to apply existing domain adaptation algorithms to our fisheries data we realized there were serious limitations in how existing algorithms were trained and evaluated. We were able to develop a new domain adaptation framework, published earlier this year in Transactions on Machine Learning Research, that addressed these limitations and led to advancements in fish counting, and even self-driving and spacecraft analysis.

One line of work that I’m particularly excited about is understanding how to better develop and analyze the performance of predictive ML algorithms in the context of what they are actually used for. Usually, the outputs from some computer vision algorithm — say, bounding boxes around animals in images — are not actually the thing that people care about, but rather a means to an end to answer a larger problem — say, what species live here, and how is that changing over time? We have been working on methods to analyze predictive performance in this context and reconsider the ways that we input human expertise into ML systems with this in mind. CODA was one example of this, where we showed that we could actually consider the ML models themselves as fixed and build a statistical framework to understand their performance very efficiently. We have been working recently on similar integrated analyses combining ML predictions with multi-stage prediction pipelines, as well as ecological statistical models. 

The natural world is changing at unprecedented rates and scales, and being able to quickly move from scientific hypotheses or management questions to data-driven answers is more important than ever for protecting ecosystems and the communities that depend on them. Advancements in AI can play an important role, but we need to think critically about the ways that we design, train, and evaluate algorithms in the context of these very real challenges.

By stimulating cancer cells to produce a molecule that activates a signaling pathway in nearby immune cells, MIT researchers have found a way to force tumors to trigger their own destruction.

Activating this signaling pathway, known as the cGAS-STING pathway, worked even better when combined with existing immunotherapy drugs known as checkpoint blockade inhibitors, in a study of mice. That dual treatment was successfully able to control tumor growth.

The researchers turned on the cGAS-STING pathway in immune cells using messenger RNA delivered to cancer cells. This approach may avoid the side effects of delivering large doses of a STING activator, and takes advantage of a natural process in the body. This could make it easier to develop a treatment for use in patients, the researchers say.

“Our approach harnesses the tumor’s own machinery to produce immune-stimulating molecules, creating a powerful antitumor response,” says Natalie Artzi, a principal research scientist at MIT’s Institute for Medical Engineering and Science, an associate professor of medicine at Harvard Medical School, a core faculty member at the Wyss Institute for Biologically Inspired Engineering at Harvard, and the senior author of the study.

“By increasing cGAS levels inside cancer cells, we can enhance delivery efficiency — compared to targeting the more scarce immune cells in the tumor microenvironment — and stimulate the natural production of cGAMP, which then activates immune cells locally,” she says. “This strategy not only strengthens antitumor immunity but also reduces the toxicity associated with direct STING agonist delivery, bringing us closer to safer and more effective cancer immunotherapies.”

Alexander Cryer, a visiting scholar at IMES, is the lead author of the paper, which appears this week in the Proceedings of the National Academy of Sciences.

Immune activation

STING (short for stimulator of interferon genes) is a protein that helps to trigger immune responses. When STING is activated, it turns on a pathway that initiates production of type one interferons, which are cytokines that stimulate immune cells.

Many research groups, including Artzi’s, have explored the possibility of artificially stimulating this pathway with molecules called STING agonists, which could help immune cells to recognize and attack tumor cells. This approach has worked well in animal models, but it has had limited success in clinical trials, in part because the required doses can cause harmful side effects.

While working on a project exploring new ways to deliver STING agonists, Cryer became intrigued when he learned from previous work that cancer cells can produce a STING activator known as cGAMP. The cells then secrete cGAMP, which can activate nearby immune cells.

“Part of my philosophy of science is that I really enjoy using endogenous processes that the body already has, and trying to utilize them in a slightly different context. Evolution has done all the hard work. We just need to figure out how push it in a different direction,” Cryer says. “Once I saw that cancer cells produce this molecule, I thought: Maybe there’s a way to take this process and supercharge it.”

Within cells, the production of cGAMP is catalyzed by an enzyme called cGAS. To get tumor cells to activate STING in immune cells, the researchers devised a way to deliver messenger RNA that encodes cGAS. When this enzyme detects double-stranded DNA in the cell body, which can be a sign of either infection or cancer-induced damage, it begins producing cGAMP.

“It just so happens that cancer cells, because they’re dividing so fast and not particularly accurately, tend to have more double-stranded DNA fragments than healthy cells,” Cryer says.

The tumor cells then release cGAMP into tumor microenvironment, where it can be taken up by neighboring immune cells and activate their STING pathway.

Targeting tumors

Using a mouse model of melanoma, the researchers evaluated their new strategy’s potential to kill cancer cells. They injected mRNA encoding cGAS, encapsulated in lipid nanoparticles, into tumors. One group of mice received this treatment alone, while another received a checkpoint blockade inhibitor, and a third received both treatments.

Given on their own, cGAS and the checkpoint inhibitor each significantly slowed tumor growth. However, the best results were seen in the mice that received both treatments. In that group, tumors were completely eradicated in 30 percent of the mice, while none of the tumors were fully eliminated in the groups that received just one treatment.

An analysis of the immune response showed that the mRNA treatment stimulated production of interferon as well as many other immune signaling molecules. A variety of immune cells, including macrophages and dendritic cells, were activated. These cells help to stimulate T cells, which can then destroy cancer cells.

The researchers were able to elicit these responses with just a small dose of cancer-cell-produced cGAMP, which could help to overcome one of the potential obstacles to using cGAMP on its own as therapy: Large doses are required to stimulate an immune response, and these doses can lead to widespread inflammation, tissue damage, and autoimmune reactions. When injected on its own, cGAMP tends to spread through the body and is rapidly cleared from the tumor, while in this study, the mRNA nanoparticles and cGAMP remained at the tumor site.

“The side effects of this class of molecule can be pretty severe, and one of the potential advantages of our approach is that you’re able to potentially subvert some toxicity that you might see if you’re giving the free molecules,” Cryer says.

The researchers now hope to work on adapting the delivery system so that it could be given as a systemic injection, rather than injecting it into the tumor. They also plan to test the mRNA therapy in combination with chemotherapy drugs or radiotherapy that damage DNA, which could make the therapy even more effective because there could be even more double-stranded DNA available to help activate the synthesis of cGAMP.