I’m going to remind you of something you already knew: nature is fascinating. Who hasn’t been amazed by the synchronized flight of starlings? Or by the skilled spiders when they spin their webs? No less impressive are the dances of bees to communicate with each other when they find a source of food. But what is really fascinating is not the ability of these animals to perform these behaviors. Starlings don’t need to spend hours in the library to learn to coordinate with their own kind, nor do spiders need to watch sewing tutorials, nor do bees need dance lessons. What’s really fascinating is our ineptitude! But are we really inept? No! That’s what I try to tell myself when for the fourth year in a row I play the same Chopin waltz on the piano, each time achieving an improvement imperceptible to the human ear. Both spider web weaving and virtuoso piano playing are behaviors bordering on divinity. However, the former is an innate behavior and therefore needs no conscious learning and the latter - unfortunately - is an acquired behavior that needs years and years of practice. For an innate behavior to occur and be repeated generation after generation, the brain of each individual must follow the same set of recipes. So, although each brain is a world and the connections between neurons vary between individuals, there is also a basic wiring that reminds us that we are also the same. I could go on writing about how amazing the piano is, but you’ve come this far to learn something about brain wiring and about science. So let’s focus a bit with Brian.
Brian Chen is an Associate Professor in Neuroscience at McGill University. He started his lab at McGill University in 2009. Chen performed his graduate work with Dr. Karel Svoboda at Cold Spring Harbor Laboratory and then worked with Dr. Dietmar Schmucker at the Dana-Farber Cancer Institute and then Dr. Josh Sanes at Harvard University for his postdoctoral research.
Juan García Ruiz: What brought you to do research?
Brian Chen: As a child, I admired how ancient philosophers spent much of their time thinking and trying to answer deep questions about how the world works. Eventually I learned that modern science was the same pursuit, where I could spend much of my time thinking and trying to answer deep questions. I had a love of nature and the outdoors growing up, and was fascinated by the natural world, in particular animal behavior. As an undergraduate student, I sought to study animal behavior and this led me to neurobiology. There, I did research on the auditory system of bats. As a graduate student, I studied experience-dependent plasticity in rodents, specifically, how anatomical changes in neurons accompany electrical changes in their receptive fields. I had so much fun doing research as a student that I decided to pursue it as a career path.
JGR: What is your research about?
BC: The question that drives me the most is how an animal can perform complex behaviors without any learning or prior experience. There are so many amazing and beautiful things that the animal kingdom can do without learning, from our own simple pain reflexes to very complex insect behaviors that are performed as soon as the animal is born. In other words, how are the hard-wired neural circuits that underlie these innate behaviours wired up?
Obviously, all of the instructions required to wire up these neural circuits have to be embedded somehow in the animal’s genome. My goal is the decipher these molecular instructions to understand how to go from molecules to fully connected and functioning neural circuit. Right now, I am trying to identify a list of molecules that are necessary and sufficient to wire up a neural circuit in the fruit fly Drosophila melanogaster. If I can identify a list of molecules that I can take and put into another neuron and re-wire it to change an animal’s perception and behavior, then I feel I would be on the first step to understanding how to wire up a neural circuit. Yes, I am starting with a single neuron, before moving to a full neural circuit, which is far from an entire brain, but it is a difficult problem. As I said, I like answering deep questions.
JGR: When did this project start and what have been the main discoveries you have made?
BC: This project started as soon as I started my lab in 2009. The main discoveries so far are lists and lists of molecules involved in and required for wiring up a particular neuron that we study, as well as a better understanding of how some of these molecules and this neuron works. Some of this work on characterizing these molecules and the neuron has been published, while other molecules we are soon to publish along with our lists.
JGR: How do you study neural circuits, in a nutshell?
BC: We mostly use the fruit fly Drosophila melanogaster as an animal model to study neural circuits. We use Drosophila because of its advantages in genetics from the past 100 years as a genetic model organism, and its track record as an exceptional model system for studying neural circuits. The main techniques we use are high-resolution imaging of single neurons, single cell RNA sequencing (so we can identify molecules used for wiring inside a single neuron), and then lots of other clever molecular biology and molecular genetics tricks that Drosophila allows for, or makes easier, like CRISPR-Cas9 genome editing or single neuron genetic knockouts. We occasionally investigate how these molecules function in the context of the mammalian brain using mice and also human neurons from stem cells in cell culture.
JGR: To which extent we can learn about our wiring from animal models?
BC: Over 60% of the Drosophila genome is homologous to the human genome, and much of the pioneering work on identifying and characterizing brain wiring molecules was performed using Drosophila melanogaster as an animal model. We would understand very little about our own brain wiring without the use of animal models throughout the long history of neurobiology. Classic experiments using models as Drosophila melanogaster, grasshoppers, chicken embryos, zebrafish, frog tadpoles, rats, and mice, produced ground-breaking science that revealed fundamental principles underlying how brains wire up that would not have been possible without animal models.
And that would be my simplistic answer to your question, but I appreciate that this is a particularly insightful question, because it is specifically about brain wiring and not just brain function. Figuring out how the human brain works, the most complicated machine in the known universe, is already hard enough without animal models. Right? Because human brains are all but off limits to invasive research—you don’t have easy access to a human brain, and I’m not just talking about the skull, you can’t take samples of living brain tissue either, you can’t perform whole cell patch-clamp recordings in a human brain, you can’t delete one area and over-excite another and see what happens, you can’t just root around inside someone’s brain, because all of those things are unethical because our brains are who we are. So it’s hard enough studying how the human brain functions without the use of animal models.
But, you asked about our brain wiring. A complete contrarian could say that all knowledge about how the human brain is wired that comes from animal models are completely irrelevant and useless, because they come from animals that are not Homo sapiens. Thus, we can only use human brains to study human brains. Or you can say, well, regardless of whether animal models are relevant to human brain wiring or not, I will only use human brains. From there you can basically only use human genetic association studies, where variations or mutations in genes cause wiring phenotypes in humans with diseases, disorders, or differences that manifest visibly, in other words, very obvious clinical signs. Beyond that, again since we are talking about how brains are wired, and this means molecules, you can only manipulate these molecules in human neurons from stem cells in a dish and then see what happens. At best you could manipulate human neurons in a tissue organoid in a dish, and that’s pretty much it. And that’s it, there’s nothing beyond that. No behavior, no natural or complex circuitry, no functional output from a brain. So essentially, what I’m saying is that there is no other way!
I like this question, it’s a great question because it made me think about the unique problems associated not just with doing neurobiology, but in particular with molecular neurobiology.
JGR: Why is it relevant to understand nervous system wiring?
BC: Understanding how nervous systems in general are wired is important and relevant because it deepens our understanding of the animal kingdom, and the how and the why of animal behavior, including for humans. For example, understanding how genes control wiring is relevant to human disabilities associated with brain wiring such as autism or schizophrenia, and for re-wiring of damaged nervous tissues. Thus, future applications could include enhancing the neural capabilities of those with mental disabilities, or recovering from damage to the nervous system.
JGR: Understanding of the instructions of nervous system wiring is a tricky task, but do we have the knowledge and means to go from understanding to build a brain or for the moment it’s still an ambitious aim?
BC: It is definitely an ambitious aim at the moment! The main obstacles are the vast complexity of genomes, whether it is Drosophila melanogaster or human, and the vast complexity of neural circuits. So for example, in a single neuron you could have thousands of different molecules working to allow the neuron to function properly and that comes from the complexity of the genome, with thousands of genes being active at any given moment in time, and then even in a simple neural circuit in an animal you can have fifty neurons and thousands of connections amongst them.
JGR: You have created a bioinformatics database. What is GeneDig?
BC: GeneDig is a web application that I developed for easy and efficient access to genomics data and analysis. As I mentioned, genomes are very complex because they are large (at billions of letter codes), they contain DNA and RNA and protein information, and there are lots of genomes that are sequenced, from coronaviruses, to bacteria, fungi, plants, animals, humans, and individual humans as well. This makes it very complicated to understand or even access the most basic genomic information, like what genes are involved in a certain disease and what are the DNA, RNA, and protein sequences for that gene?
As a postdoc, I found it frustrating to use the publicly available, ridiculously complex databases just to get the RNA sequences for a gene to do some basic experiments. Once I figured out how to do it, once I started my own lab, I still had to train people each time on how to do what should be easy, basic, bioinformatics tasks. Like, let’s say I have a patient mutation in a disease gene that I would like to recreate in the lab to investigate. Finding the sequence of a gene using the public databases is super hard for a high-school or undergraduate. Then, you have to locate the site of the mutation, then find the RNA sequence, and then how the DNA and the RNA and the protein all line up together to design your experiments to recreate the mutation. Trying to convert this into the mouse or fly homolog is even harder! Similarly, doing any protein structure-function work is equally painful, going from amino acid to RNA, let alone how that relates to the genomic DNA. Those examples are each about 20 bioinformatics steps that use about 20 different databases and websites, that each take half an hour to an hour to resolve, but takes GeneDig a few seconds.
Anyways, so I created GeneDig to solve all of these problems. Back then I realized that meaningful access to genomic and bioinformatics information was super important and only going to get more important in the future. So, my goal with GeneDig is to make all genomic information easily accessible and useful. I downloaded all of the publicly available sequences onto my servers, all of the disease information, all of the chromosome information, everything I could find. Then from GeneDig, you can simply type in any gene or disease into the search box, or change the organism to any sequenced organism to search their genome. Or you can start browsing the chromosomes instantly by just clicking on the name of the organism. Once you start browsing the genome you can see the relationship between disease mutations, DNA sequence, RNA sequence, protein amino acid sequence, and protein domain altogether.
GeneDig isn’t really meant for the general public even though I built it with that in mind as a guiding principle. Fortunately, it is often used across the world for biology education, like in Brazil, Indonesia, and India, so I must be doing something right. Unfortunately, I don’t have a team working on it, so it’s super hard to fix things and roll out new features that I’ve had in mind.
JGR: How can researchers benefit from GeneDig?
BC: I built GeneDig from the ground up with the average user, say at the high school to undergraduate level, in mind. In GeneDig, the relationships between genome DNA and RNA and protein amino acid and protein structure are all bound together. If you highlight any sequence in, say for example, RNA, then the corresponding DNA and protein gets co-highlighted. By the way, this also makes it much easier to do CRISPR-Cas9 genome editing when you are starting from a protein domain or RNA sequence! Obviously, this saves a ton of time because these relationships between DNA, RNA, and protein are simultaneous, and now you don’t have to have lots of different websites open at once, plus a text file open where you copy and paste sequences back and forth between websites, and then line things up in your text documents, and all that. Everything just takes a few seconds and you see it in real time.
My lab uses all the time for finding coding sequences of RNA, the untranslated regions of RNA, CRISPR-Cas9 genome editing of human, mouse, and fly genomes, and finding where to mutate RNA starting from an amino acid sequence. It’s a real time saver. We are also using it to store all of our DNA, RNA, and protein sequences online. We also play around with GeneDig’s automated molecular cloning algorithm where it automatically tells you how to perform molecular biology. The algorithm spits back the protocol for how to do a PCR assembly, Gibson cloning, restriction digests, and even Golden Gate assembly. So that’s been very useful, too. I hope with some more tweaks, the algorithm will be the end of all human input into molecular biology and molecular cloning. No more human mistakes!
JGR: What do you like the most about research?
BC: I definitely have the best job in the world! I get to investigate the most complex machine in the known universe and unravel the deepest mysteries of biology. I am definitely very lucky to get to do what I do. I have no boss, I have complete job security, and complete academic freedom, so I can pursue fun things like GeneDig and other exciting projects. I get to lead and interact with a phenomenal team of really smart undergraduate students, graduate students, medical students, postdoctoral fellows, engineers, computer programmers, and technicians. Every day there is something new for me to do and learn about and I am constantly being challenged intellectually and creatively, so what’s not to like? On top of that, my colleagues here at McGill are very smart, nice, a lot of fun, and down to earth, very Canadian.
I like to be in lab doing things, so I prioritize time for that. I enjoy meeting with my lab discussing their latest progress and results, doing experiments, performing surgeries and dissections, doing molecular biology, imaging, building new equipment and machines, building new software, and just playing around in general. I balance the projects in the lab in the following order of priority: curiosity-driven and interesting to me (e.g., how hard-wired circuits are wired up), interesting or meaningful to society (e.g., how a disease gene alters a neuron’s function), societal utility (e.g., our high-throughput drug screens to identify drugs that alleviate a disease), and inherent beauty (e.g., making an auto-luminescent fruit fly). Thus, the majority of the projects in my lab are driven by curiosity and having fun.
One of my most favorite things about research is the exciting moment when you realize you are the first person to observe your discovery. I felt like that many times in my career and I enjoyed and cherished each time the exciting combination of wonderment and pioneering spirit.
JGR: What have you learned by doing research?
BC: I’ve learned a lot about working with others. I’ve learned how to be a leader and a follower, and how to communicate effectively and how to have difficult conversations. I’ve learned time-management, how to get things done, and how to balance detail and precision with the big picture, and how to handle set-backs and delivering results. I’ve also learned a lot about teaching in classes, small groups, one-on-one, in labs, and how to mentor different levels of people, from kids to new professors. I’ve learned how to motivate people, and how to protect people from their impulses. Doing research teaches you much more than just science and how to perform research. But, I’ve also learned a lot of new science, old science, math, engineering, programming, and of course new discoveries!
JGR: What are your thoughts about publication pressure in research?
BC: As you can see from my creation of GeneDig and some of my blog posts, I feel very strongly about free access to knowledge. So much so that I tried to create my own solution to the problem. I envisioned what scientific knowledge should look like if it were started from scratch now, without the burdens of the past, the legacy of paper publications, and the publishing companies, versus in the modern era with the internet, rapid connectivity and high bandwidth, and the examples already set by YouTube, Wikipedia, and Reddit. I thought it would look very similar to a mash-up of Wikipedia, YouTube, and Instagram, basically. Each lab would have their own channel and each person as well. Each project would have a dedicated page that gets updated in real-time. Projects can be public or privately visible. Importantly, all primary data can be stored (finally!) in cloud servers (perhaps verified through a blockchain), as well as data analysis and meta-analysis. Everything is publicly available for everyone in real time. This would standardize experiments across labs and countries, make science more transparent and collaborative, reduce fraud, overlap, and time wasted on repeating projects and experiments with negative results.
Proprietary projects and knowledge can still remain private, and peer review can still come from the community, constantly, in real time, and eternally. All scientific knowledge would finally be democratized. Good science would still be rewarded, even if it only recognizable decades or centuries after the fact, which is the case nowadays anyways. Scientific interest would stand on its own without judgement of impact, like a Wikipedia page. I’m not sure how job hunting and promotions would be handled in this scenario, but it’s a pipe dream anyways. I created a version of the web application that I called Voxfer.org, but then realized I don’t have enough time and resources to pursue my vision for pure scientific knowledge freedom.
I still think it’s crazy that scientists fret so much over a few letters in the citation of their work, like “Anonymous, et al., Generic Journal, 2021, Important discovery.” versus “Anonymous et al., Different Journal, 2021, Important discovery.” versus “Anonymous et al., Vanity Journal, 2021, Important discovery.” I don’t see how it makes any difference in the content of the manuscript. 99% of people that access scientific papers don’t care about which journal a study is published in, because 99% of scientists are the assistants, technicians, young students, undergraduates, journalists, graduate students, and postdocs, who vastly outnumber the small number of people that do care. I think modern scientists just find the work they are interested in reading, judge the science on its own merits perhaps in a journal club, and don’t pay attention to the journal that it is in because it is all online. Even if it were published in my local PennySaver magazine, I wouldn’t care, as long as it is interesting to me and the science is good (“Anonymous et al., PennySaver, 2021, Important discovery!”). Unfortunately, I also understand it’s a game we all have to play for now. It’s only a matter of time before it all becomes free and transparent though!
JGR: What would you tell future researchers to improve research quality?
BC: Learn from good scientists, ask a lot of questions, and listen a lot. Always assume you don’t know much, and always be willing to learn. Arrogance is the antithesis of science, and once you become arrogant you stop learning. Read lots of different articles on lots of different subjects. Attend lots of seminars on various subjects.
One helpful way to grow is when you are attending seminars in your field. Write down lots of questions you would ask during the seminar, and see if they line up with what others ask during and after the seminar. Discuss the seminar with others. Then see if your questions at other future seminars begin to evolve, from not understanding the material, to curious tangents, to nit-picky details (which are still important), to critical details, to critical experiments with the bigger picture in mind. Look to see whether you can differentiate for yourself whether the questions that you are asking require either a different interpretation of the data, or a different data analysis, or a different experiment entirely. Why is this helpful? In your own science you will then begin to see what the payoff is for your different experiments, and how they affect the big picture outcome. This will help you prioritize your time, energy, and money and allow you to deliver on the more important experiments. In the theoretical world, all experiments are important and every control is important, but in the real world not all experiments are equal, and some are more important than others. Find out which ones will tell you which things about your hypothesis. Find out during seminars which controls and experiments tell which things, and you will be on your way to improving your research quality because you will be thinking deeply about your science.
It is definitely important to be very critical, and to learn a lot about mistakes, fallacies, and sloppiness in science. But I feel that there is no need to emphasize that, because most scientists are already nit-picky and hyper-critical, and some tend to take pride in being negative and destructive. It is very instructive to learn from the hyper-critical though, it is still tremendously helpful.
Here are some more practical tips: This is hard to do, but make sure to continually iterate from experiment to analysis, back to experiment, and let your analysis inform the next round of experiments. Don’t wait to pile up a bunch of experimental data before you go over it. I know that lots of scientists love the experimental side of things, and find comfort in the process of doing things with your hands, I certainly do, too. But don’t use that as a procrastination method to data analysis. I love this part as well and that’s often where the real moment of discovery occurs and it is exciting. You should be putting in at least as many hours in data analysis for every hour of experiment. Stare at primary data a lot. And I mean a lot. Obsess over it. I still do, and I no longer generate much of the primary data in my lab. Let the data and the science tell you what the biology is.
Here are different exercises we do in my lab a lot. Always know what the purpose of each experiment and control is for, and specifically, what the best case outcome you expect is, the worst case outcome, and the most likely outcome. Plan ahead what you will do in each of these scenarios. Keep in mind that the worst case outcome of an experiments is often not a negative result but a uninterpretable or middling result. Those are the worst. But, if you know ahead of time what each outcome is supposed to tell you about your hypothesis, then you will much more deeply understand your projects. The great thing about this exercise is that it is all just thought experiments that get you better at your science, and they allow you to think multiple steps ahead for your projects. I also have my lab do this for planning everything in life, and thinking about outcomes and what your reaction and next steps will be given different outcomes. It is a very useful and easy habit to pick up.
Another exercise that I do is in lab or in our journal clubs, we examine a project or experiment and start with infinite time, money, and personnel. Then think about all of the possible ideal experiments and controls that you would do if you had infinite time, money, and personnel. The most important thing, again, is to list what each experiment or control will tell you about your hypothesis. This is fun because you can also start to go crazy dreaming about what you would do with lots of money. Still, you also begin to see that there are some points of diminishing returns and that, again, some experiments are more important than others. You will really begin to see what the rationale is for different experiments are, what your holes are in your experiments and how big of a hole it is, or in a journal club paper, and how it ultimately affects the main hypothesis. You can also see more clearly why you may not need a multi-million dollar clinical trial to prove something reasonably.
You will also more easily see how projects can branch off too much and not answer a central question, even with infinite resources. The end of the exercise comes with picking which experiments to do in the lab with the real world constraints of limited time, money, and personnel. Finally, the most painful part of the exercise is estimating realistically how long each experiment will take, and the multiple resources and steps and sub-steps it will take to get it done, and the reality sets in that science takes a lot of time. Nevertheless, this is a fun exercise because it allows you to be very creative and they are just thought experiments. Who doesn’t like dreaming of infinite funds?
JGR: Would you like to share a general message to the readers?
BC: Most importantly, thank you for reading this or having any interest in my science. If I have helped just one person with my long answers, even subconsciously, then this is more than worth it.
My general message to any readers is to not try and be a good scientist, but a good person. I’m not trying to be generic, so what I mean is, learn as much as you can and as many skills as you can that can be useful to society or somebody. (It doesn’t matter what the skills and knowledge are to start with). Get extremely good at a few things, and pretty good in lots and lots of other things. Then, go use these skills or knowledge to help someone. You will be pleased with the results.