Sir Henry Dale Fellowtim.firstname.lastname@example.org
Tim Vogels studied physics at Technische Universität Berlin and neuroscience at Brandeis University as a Fulbright Scholar. He received his PhD in 2007 in the laboratory of Larry Abbott. After a postdoctoral stay with Rafael Yuste at Columbia University, he became a Marie Curie Reintegration Fellow in the laboratory of Wulfram Gerstner at the École Polytechnique Fédérale de Lausanne (EPFL). Tim was awarded the Bernstein Award for Computational Neuroscience in 2012.
As a computational neuroscientist, Tim builds conceptual models to understand the fundamentals of neural systems at the cellular level. His research group is funded by a Sir Henry Dale Fellowship of the Wellcome Trust and the Royal Society.
How does your work differ from the existing research groups at the CNCB?
First of all, I am a computational neuroscientist, so I don’t do experiments, I don’t have a wet lab, and most of my time is spent in front of a computer. And then, more fundamentally, I am interested in generic neural systems, rather than the specifics of the neural networks of fruit flies. The work by other groups here at the CNCB is focused—very specifically—on studying the fly brain and then to extrapolate and understand neural systems in general. I come from the opposite direction, starting with general mechanisms that can be used to explain certain aspects of specific systems. You could say I explore the characteristics that are common to neural systems in all animals, whether fly, or human.
My work is based on building abstract models of neural networks based on a few of their essential characteristics. I am modeling how a circuit with certain qualities will behave, or respond, in a given situation.
How does theory relate to experiment?
It is probably true to say that most experimentalists study something very specific but hope to explain a general quality of neural systems. I hope my approach will be an inspiration for how systems could work and invite experimentalists to ask wide-ranging questions. A theoretical approach means you can explore any strand, or link, to generate a testable prediction which in turn can lead to discovery or a new form of thinking. While a lot of what I am doing is physically possible within the constraints of reality, those constraints are often broad. For example, if I have 10,000 cells at my disposal, how can I communicate any sort of information between them? Finding an answer to this could open up new questions when, for example, you are studying the fruit fly. This in turn could then lead to more realistic models.
What are the big unanswered questions in our understanding of neural networks?
To me there are many unanswered questions about how neurons behave at the cellular level. What makes us tick, how do we process information, make choices? What happens within neural networks to enable us to see, talk, feel, touch? How do we decide, for example, to ‘tune in’ to one sound over another, such as why you are listening to me rather than the hum of the air conditioning. These are big mysteries because there is so much we do not know about how information is transferred from point A to point B, and how it is transformed on the way there.
I want to zoom in very deeply into neural networks to understand the cellular mechanisms which make them work. At this level of scrutiny the animal or species is virtually irrelevant, as one is studying characteristics which are common to all systems. For example, when you study photosynthesis, you don’t care whether you are looking at a palm or an oak; your quest is to find the common mechanisms in all plant systems which enable photosynthesis to work.
What are some examples of where theory has advanced neuroscience significantly?
I am not sure that there is a good answer to this question, because you’re bound to forget someone. What are the ten most significant experiments ever done in neuroscience?
People usually reference Hodgkin and Huxley in response to the question of important theory work. Their paper explains the interplay of sodium and potassium channels in the generation of action potentials and is certainly a significant contribution. But theoretical and computational neuroscience per se is much younger than that paper. Hodgkin and Huxley were both experimentalists first, and theorists on the side, so to speak.
In the more recent past, say the last 20 years, it’s much more difficult to name similarly significant contributions. That is also because as a field, we still depend very much on experimental verification, we are not as self-sufficient as say theoretical physics or math. So we are often still waiting for experimental verification to know if a theoretical idea was good. A good example for this is a paper by Bienenstock and colleagues from 1982 in which they show that excitatory synaptic plasticity must be regulated somehow to avoid run-away feed back loops that would eventually lead to disfunction. The precise mechanism they proposed has not been experimentally verified, but is it a significant contribution? The paper proposes answers, creates more questions, and generally makes you think, so I would say it’s significant. My personal favourite, and this is clearly very subjective, is the prediction of a balance of excitation and inhibition on a single-cell level by Carl van Vreeswijk and Haim Sompolinsky in 1996 to explain the asynchronous and irregular firing behaviour of cortical neurons. I think that paper really influenced a lot of work that followed.
So I think once you start looking, there are many many theory papers that make a contribution. How big that contribution is probably depends on the eye of the beholder, and the subfield of the beholder as well.
There is a growing trend for combining experimentalists with theorists in research. Do you see this as beneficial?
This is a rapidly growing, but still quite small, aspect of research in neuroscience, and there are still not many institutes combining both experimentation and pure theory. But it’s a growing and exciting trend, and it is often beneficial for both sides. Communication and interplay between two parties that usually approach the same problem with a radically different set of tools and frameworks will make both parties think about these problems in a different way. And the only real way to test a theoretical idea is still an experiment. What could be better than to have an experimentalist down the hall? Vice versa, it is often very time consuming to test yet another hypothesis in an experiment. The computational neuroscientist can help to test and refine the hypothesis and sharpen the expectation of an outcome.
What are your plans for your research group?
To build a team, establish a groundbreaking neurotheory group and add to the growing lustre of a respected neuroscience institute, that would be pretty cool, no?