“What really interested me, and still interests me, is how neural activity gives rise to cognition.”
Antonio Lara, Neuroscience PhD Program alum (entering class of 2004)
Antonio Lara is the director of neuroscience at Kernel and an alum of the Berkeley Neuroscience PhD Program, but neuroscience wasn’t always on his radar. In his last year as a physics major at the University of Texas at Austin, Lara took a neurobiology course to fulfill an elective requirement and quickly became hooked on the brain. So much so, that after completing his BS in physics, he stayed another two years to earn a second BS in neurobiology.
Lara went on to do his PhD in the Wallis lab, where he studied mechanisms underlying working memory. As Wallis’ first graduate student, Lara also helped set up her lab, and gained a wide variety of skills in the process. After a postdoc at Columbia, Lara decided to go into industry, where his hands-on experience building systems from scratch became an important asset at Kernel, which was an early stage startup company at the time.
At Kernel, Lara is developing noninvasive, wearable devices to record brain activity as people move around, interact, and go about their daily lives at home or outdoors — not just in a lab or hospital setting. He is excited by the prospect that the new technology could advance our understanding of the human brain by enabling neuroscience researchers to include many more — and more diverse — participants in their studies.
Read our Q&A with Lara to learn more about his career path, how his experiences and connections at Berkeley helped him along that path, and what he enjoys about working in industry. This interview has been edited for length and clarity.
Rachel Henderson: How did you become interested in neuroscience?
Antonio Lara: It was during my last year of my bachelor’s [degree in physics]. I still needed to take a few elective courses, and just for the fun of it, I took a neurobiology course. I really, really enjoyed it. The next semester I took another neurobiology class because I liked it so much.
By the time I was finished with all of my physics, I realized that neuroscience was way more interesting to me than physics. So I finished up all my physics requirements, graduated, and asked if I could stay and do another bachelor’s. They said yes, they agreed for some reason. Then I did a bachelor’s degree in neurobiology and really got into it.
The thing I remember the most was a lab course where we learned how to patch clamp [Ed. note: a technique used to record current from individual cells]. It was so cool to me to see action potentials happening right in front of me. And I was the one that put the electrode in the neuron and was recording all of this data. That was a pretty cool feeling, so after that I was hooked.
RH: Tell me about your general experience in the Berkeley Neuroscience PhD Program.
AL: I think it was challenging, but I also really liked that. I did my PhD with Joni Wallis. What really interested me, and still interests me, is how neural activity gives rise to cognition. At first I found that very challenging, it’s a lot of work. I didn’t really realize how much work it is. Getting neural data recordings is just very time-consuming and challenging because you spent months training these [animals] to do whatever cognitive tasks you want them to do.
I liked the [lab] rotation system a lot. I liked being exposed to very different techniques, questions being asked, and how do you answer those questions. I was also really interested in the computational part of [neuroscience]. I think when I started, the Redwood Center [for Theoretical Neuroscience] was just getting off the ground, and I used to attend their seminars and take their classes. I really enjoyed that a lot.
I became really good friends with a core group of people — a couple of guys above my year [in the PhD program] and a few guys from my year. We used to hang out a lot. I still miss Berkeley, [especially] the outdoor activities that you can do there. I really got into sailing my last few years at Berkeley and I miss that a lot. I used to live close to the hills, and I’d go biking back there, and trail running. After Berkeley, I moved to New York, and it was a huge difference. I really missed that from Berkeley.
RH: What was your thesis project about?
AL: The overarching theme of my thesis was trying to figure out the mechanisms that give rise to working memory [in the frontal cortex]. It was kind of split into two big projects. One of them was about working memory for taste. The idea was to figure out if kind of the same mechanisms that people have studied for visual stimuli, for example, also held up for other modalities like taste.
The second [project] was still in the working memory domain, but investigating the limits of working memory. There’s this very famous paper from a while ago that said that people have a limit of about seven things — they can keep seven things in working memory at one time, plus or minus a few. There’s been some debate and people say now it’s four or people say it’s more, but the whole idea of the second part of my thesis was to figure out — what are the mechanisms in the brain that give rise to that limit?
RH: What did you do after you graduated?
AL: I stayed for two more years with Joni, finishing up. And I got really interested in new analysis techniques that were coming out from the motor control literature that were looking at neural activity in a different way than I was used to looking at it. I started seeking out a postdoc with someone who did that sort of work. I ended up at Columbia in the lab of Mark Churchland who was one of the pioneers of this new approach. My idea was: I want to go and work with Mark and learn all of these tools, all of these new approaches, and then eventually get my own lab and come back to cognition. Because all these tools really started in the motor control field, and I wanted to apply them for cognition.
I did a postdoc in Mark’s lab; I loved it. I learned a ton from him and from that world. But by the end, when I was there for four years, I started looking into the academic job market and there was nothing that really attracted me or stood out to me. I was liking less and less the idea of the whole academic system where you try to find a job wherever you can. So I started looking into industry.
Just randomly, I read an article in the Washington Post and they were describing the sort of stuff that Kernel was starting to do. At the time, the only person I knew in industry was someone from Bob Knight’s lab [at Berkeley] — she used to be a postdoc. I emailed her and said, ‘You’re in industry, have you heard of this company? Are they legit?’ Just that, really out of curiosity. She said, ‘Yeah, actually. I know someone who works there — Chris, who was also at Berkeley. I’ll put you in touch with him.’ I said okay. I wasn’t really looking to get in touch with him, but sure. I just wanted to know: what he’s doing, and does he like it? And one thing led to another, and I ended up working for Kernel.
RH: What does Kernel do and what you do there?
AL: Generally, we’re trying to build hardware for interfacing with the brain, and trying to take high-quality brain recordings and make them more accessible. Not just accessible to highly specialized people with PhDs at a research institution or hospital [who] take years to get trained on how to operate these machines. We want to make it easy, so that eventually, it could even be like every person has a brain interface in their house. It could be as simple as just putting something on, and then you calibrate it and start recording brain data.
There are many things this would enable, like large-scale neuroscience studies, which is what really interests me now. Going beyond one or two, or tens [of participants]. The really large-scale studies [now] have hundreds of participants. To me, it would be amazing if we could get thousands of participants — brain recordings from thousands of people from every part of the world. A really diverse participant pool, not just university students or patient populations. Once that is possible, my hope is that we’ll start learning things that weren’t accessible before because of small populations, or the limited sampling of the whole population that we’re doing.
Part of the big, big dream [of Kernel] is that we also start incorporating things like human-to-human interaction. I know there’s a big social neuroscience field, but the idea here is to have data from loads and loads of people while they’re interacting in a social setting, outdoors, in their normal lives. Not in the lab at a research university, hospital, etc. So that’s one of the visions of Kernel.
RH: What is the hardware like?
AL: One of the devices that measures blood flow [Ed. note: as a measure of brain activity] is called fNIRS [functional near-infrared spectroscopy]. The idea is that you get a signal that’s comparable to fMRI, although not as much, because we’re limited to just the cortex [of the brain].
The second technology that we’re developing is MEG [magnetoencephalography]. We’re not the first ones to develop this, but there’s a new technology that’s been around for a few years that doesn’t require cryogenics to make the sensors work. We operate them at room temperature, and they’re relatively compact. So you can build systems equivalent to a research-grade system that are really, really small. That’s the second project that we’re working on, trying to make a MEG system that is small, relatively speaking, and it could be worn — not just when you’re sitting in a tube. [Currently for] MEG, you sit and put your head inside this [vertical] tube, just your head. And inside the tube are all the sensors. But you really can’t move at all, and this tube is huge. What we’ve developed is head-worn and you can move around with it. The ultimate idea is that you [could] even walk around and you don’t need all this crazy equipment.
RH: What are some things you like about your job, and that might be different in industry versus academia?
AL: The biggest thing for me is getting to work with a team, not just by yourself. In academia, you’re not completely by yourself. But I think the model is that you get a PhD, go into a postdoc, and eventually [become] a professor, and you’re being trained to be independent, to be able to do everything on your own. Obviously, you work with grad students and postdocs, but the whole focus is (at least this is how I perceived it) on making you an independent researcher with your independent projects, so that you can eventually go off and start your own lab.
For me, one of the biggest differences that I really enjoy about industry is that is absolutely not the case. You’re working with a team of really capable people. I’ve worked with some of the smartest people [in industry]. Not that people at Berkeley weren’t smart [laughs], but what I mean is [that members of the team have] very diverse backgrounds. Not just neuroscience, but electrical engineers, mechanical engineers, physicists, even the business development aspect of it, and regulatory [aspects]; it is such a diverse team. And we’re all working towards the same goal. That’s what I like most about industry.
RH: Did you have any experiences during your time as a PhD student at Berkeley that helped prepare you for your current career?
AL: When I started in Joni’s lab, I was her first grad student. There was a postdoc, but it was really, really early days for her lab. So I got to learn how to set up everything. I got to learn how to put systems together, how to write software to have all these different systems — like an eye tracker, the neural recording, the behavior of a computer that displays [stimuli] — talk to each other [and] synchronize. I had to build a lot of the recording equipment with my own hands. So I got to do a lot, and I got to learn a huge variety of skills.
That also kind of repeated when I got to Columbia, because I was one of Mark’s first postdocs, and helped him set up his lab from scratch as well. When I got to Kernel, it was very helpful to be able to know how to build systems from the ground up. Because in the early days, that’s what we did a lot [since we] were an early stage startup. We’d have to do everything ourselves, and I would say that everything I learned at Berkeley and Columbia really helped.