Leia este artigo em português.
In 2025, the International Brain Lab (IBL) published a landmark Nature paper detailing a large-scale, brain-wide map of neuronal activity in mice. The map—which features more than 600,000 neurons, recorded across 12 different labs, in animals performing the same behavioral task—provides one of the most complete assessments of neuronal computation in the brain.
When Uruguayan researchers Matias Cavelli and Joaquin Gonzalez looked at the dataset, they thought it could benefit from the addition of a major physiological variable they’ve been studying over the past few years—breathing. Adding it to the IBL dataset would help researchers better understand how respiration modulates decision-making or perception, says Cavelli, assistant professor at the Universidad de la República.
To make that possible, Cavelli and Gonzalez are developing computational tools to decode respiration from mouse orofacial video recordings collected during the IBL’s behavioral task. “Building on this data to decode breathing from video would allow the scientific community to investigate important questions about how breathing directly impacts neural function in different contexts,” says Vani Pariyadath, IBL’s executive director. “This line of work will also enhance tools for removing breathing-related confounds in neural data. So this project is expected to have broad impact.”
The effort is one of seven inaugural IBL Partner Projects joining the consortium of 22 labs this year. The projects all build on IBL datasets with technical support from IBL engineers and research infrastructure. “I hope this will help establish my lab—and Uruguay more broadly—as a regional hub for open neuroscience,” Cavelli says.
Beyond the scientific findings their tools might help enable, the pair hope that their work can inspire other researchers in Latin America to embrace open neuroscience. “The economic and geographical barriers we face as a continent force us to think about science from this perspective,” says Gonzalez, a postdoctoral fellow in Gyorgy Buzsaki’s lab at New York University. “I want to go back to Uruguay to have my own lab like Matias. Open neuroscience projects like this make me believe it’s possible.”
The Transmitter spoke with Cavelli and Gonzalez about how they developed the project, what they hope to learn and why Latin American neuroscientists need to embrace open neuroscience.
This interview has been edited for length and clarity. It was conducted in Spanish and translated to English by Francisco Rivera Rosario.
The Transmitter: How did you get started with the project? How did you get involved with the International Brain Lab?
Matias Cavelli: In the past few years, Joaquin and I have been part of a group of researchers that have been really interested in how the sensorimotor act of breathing affects behavior and brain-wide neuronal activity. It has become increasingly clear that nasal and oral breathing modulate activity almost everywhere in the brain. For example, the breathing cycle can modulate a variety of neuronal interactions in the neocortex and hippocampus, some of which we know play a fundamental role in memory consolidation during sleep. But there are still a lot of open questions. We thought that developing tools to couple IBL’s Brain-Wide Map with respiration would be useful, by making it possible to see how it modulates decision-making or perception. When we saw that the IBL put out a call for new partners to work with their datasets, Joaquin and I submitted a proposal for the project.
Beyond the scientific motivation, connecting with IBL through this collaboration is also very important for bringing these standardized setups and database management practices to my lab and to Uruguay as a whole.
Joaquin Gonzalez: The first person I heard talk about respiration and how it modulates brain signals was actually Matias, who was following Adriano Tort’s work in this field (a Brazilian researcher based in Natal who later became my advisor). The project arose as a kind of attempt to see if we could measure respiration through video signals. We want to develop a tool that could help us answer a series of questions that seem quite important to me, like whether the phase of respiration really modulates decision-making, and just how global this brain modulation by respiration really is.
We aim to use machine learning tools to build a model that can decode these variables from these experiments. I think in the coming years, neuroscience needs to move in this direction to pursue experiments that are alive in a sense, where you can always keep adding variables and getting new information.
TT: How far along are you in the project?
JG: We’re in the proof-of-concept phase of showing that we can actually decode respiration, meaning the respiratory waveform signal, from neural activity using a video recording set up. Our preliminary data already shows we’re basically able to do it. In general, we see that when the animal breathes, it produces changes in body posture—you can see it through chest activity, how the chest inflates and deflates, and you can see changes in whisker and nose movement. The challenge emerges when the animal changes its behavior (sniffing versus lying still, or when it falls asleep), because breathing becomes more irregular. Motor patterns appear to shift as well during these periods.
TT: What is the value of open neuroscience projects like this one for Latin America?
MC: I think it has huge potential. Through the brain-wide map and this whole open-science initiative, you can simply go analyze this data from your computer in Uruguay, Brazil, Chile, Argentina, to ask interesting questions. That lets us cut the costs of building this technology in South America, where our project funding is much lower.
Sometimes the hardest part is visualizing that it’s possible. I think that’s a challenge for neuroscience in Latin America—Joaquin and I would like to be an example and start spreading the practice and adoption of open science to other people.
JG: I’ve been thinking for years about how to build a fruitful career from Latin America, given the limitations that are involved. We need more open science projects and more fluid exchange of people and ideas within the community in the region. Having anchored ourselves in an international project, we hope to attract more people from the region to do neuroscience in Uruguay.
MC: I also think we need to shift the thinking about how science is done and help the community understand the value of open neuroscience. For instance, Joaquin published a recent paper on how gamma oscillations in the olfactory cortex arise from local feedback inhibition during breathing, working entirely with an open database. Before he submitted it, some colleagues warned him that reviewers might see it as a weakness that Joaquin had not collected the data himself. And yet, the paper ended up being highlighted by the journal and later received a Ben Barres Spotlight Award. It’s the generational shift. We need to break certain dogmas we have, that science has to be done in certain ways just because that’s historically how it’s been done.
TT: Do you have any tips for researchers in Latin America on how to develop open science partnerships?
JG: Look for them in the first place; opportunities may already be available but local communities may not be aware. Try to keep yourself updated with what is going on in the world. Don’t be discouraged if things don’t work the first or second time, keep pushing, there are local and global biases we need to overcome.
MC: My advice is to overcome our fear of collaborating openly. The questions neuroscience is trying to answer are becoming larger and more complex, making them increasingly difficult to tackle alone. We are often afraid of losing our ideas or the credit for them, but science becomes stronger when we share knowledge transparently and build trust.
TT: Is Latin America well-suited to develop open science projects?
MC: Latin America may not yet be fully prepared to develop open-science projects, but open science is particularly well-suited to our needs. It gives us access to data and resources once limited to a small number of laboratories, and we need to seize this opportunity while developing the capacity to contribute our own data and tools.
