Document Actions

You are here: Home / Research / Meet the Scientist / Interviewing Markus Diesmann and Sonja Grün

Interviewing Markus Diesmann and Sonja Grün

Forschungszentrum Jülich

Tell us how you became involved in science.

Markus Diesmann: For me, it has somehow always been clear that I wanted to become a scientist. I even told my teachers in primary school. [short break]

Sonja Grün: Yes, but what were the next steps? It is a long way from primary school to research...

Markus Diesmann: I came into contact with brain research by playing with computers in high school. After graduation, I was wondering what to study to get into this field. At first, I thought the combination of computer science and psychology might be a good basis, but then I considered to study a fundamental science first—which brought me to physics in Bochum. During my exchange year at the University of Sussex I came back to brain research, and I also chose a topic from neuroscience for my diploma thesis.

Sonja Grün: For me, it has been less obvious. I can only remember that being a child in elementary school, I once said I wanted to keep on learning for my lifetime—which in retrospect I do interpret in a way that I wanted to become a scientist. After all, that’s what you do in science! However, my way has been much less direct than Markus’ path. I started with an apprenticeship as an electronics technician at IBM. However, I did not want to repair printers all my life, so I went for the Abitur via continuation education. I first came into contact with brain research during a seminar offered by the Max Planck Institute for Biological Cybernetics in Tübingen, when I studied psychology. Although I did quit psychology to switch to study physics soon after, I returned to the MPI [comment: Max Planck Institute] during the advanced study period. Later, I prepared my diploma thesis there and fell in love with brain research.

Porträt Sonja und Markus

Markus Diesmann and Sonja Grün. Copyright: Forschungszentrum Jülich


What happened then?

Markus Diesmann:In my thesis I dealt with Moshe Abeles’ idea of synfire chains. According to this theory, there are networks in the brain built in such a way that they can support the propagation of precisely matched action potentials. Marc-Oliver Gewaltig and I wanted to model them in simulations. For this we obtained copies of Moshe’s lab book in Hebrew. Of course, this was not very comprehensible for us, but with Moshe’s help we worked through it. These were the origins of the NEST simulator, which I am still working on today. And this is also when I met Sonja.

Sonja Grün: Like I said, I completed my diploma thesis at the Max Planck Institute for Biological Cybernetics under the supervision of Ad Aertsen and Hermann Wagner. I was involved in sound localization in barn owls. We already had some experimental data. My job was to build a model that allowed to model and simulate—or understand—the data. When Ad left to go to Bochum, he offered me position as a doctoral student. After one year, Markus and Oliver came to prepare their diploma theses with him. Markus and Oli were already inseparable at the time.

Markus Diesmann: Sonja organized desks for us from the director of the institute. This was quite helpful...

Sonja Grün: And then I was given a hard time by the director because as a spoiled MPI brat, I had dared to ask for desks for diploma students...

Markus Diesmann: At the beginning, we worked on quite distant topics: Sonja on data analysis and I on abstract synfire chain problems.

Sonja Grün: Our collaborations intensified when we followed Ad to the Weizmann Institute of Science in Rehovot, Israel. That was in 1994. Together we analyzed experimental and synfire network data using an analysis method that I developed in my PhD thesis. We also became a couple in privat life at this time.

Markus Diesmann: Meanwhile, I had become a graduate student myself. However, our joint stay at the Weizmann Institute did not last long: after Sonja had finished her doctorate, she became a postdoctoral fellow in Moshe Abeles’s lab at Hebrew University in Jerusalem.

Sonja Grün: That was a very important period in my career. Using multi-electrode recordings, we studied how movement and movement planning are represented in the cortex of the monkey. This way, I got to know the whole chain from the animal to data acquisition to computer files. If I had not gained this insight, I would not be able to cooperate so well with experimenters today. And, of course, it is great to work with Moshe Abeles. He is creative, has crazy ideas and makes you think really hard.

Markus Diesmann: In 1996, I moved to Freiburg for two years when Ad was offered a professorchip at the university there. After that, I worked as a group leader at the Max-Planck-Institute for Nonlinear Dynamics and Self-Organization under the direction of Theo Geisel. Tom Tetzlaff started his diploma thesis with me, and later Abigail Morrison and Sven Goedeke joined. Substantially, I continued dealing with questions about nerve cell activity in networks. We tried to generate more and more realistic models. Besides that, the collaboration with Sonja continued. Especially the time when Sonja was also at a Max Planck Institute, the MPI for Brain Research in Frankfurt, was very productive.

Sonja Grün: We haven’t reached that part yet… When I finished my postdoc in Israel, I joined the Max Planck Institute as a senior postdoctoral fellow in 1998. There, I established my own work group in the Department of Wolf Singer. After my stay with Moshe, I quit doing experiments myself. Instead, I focused on data analysis and development of analysis methods, and collaborated with experimentalists, such as with people from the Singer lab. At the end of 2002, I obtained a research lectureship—a kind of junior professorship—by the Stifterverband für die Deutsche Wissenschaft. Randolf Menzel in Berlin had been very supportive of this and he then brought me on board of his Department of Neurobiology at Freie Universität. My group steadily grew bigger over the time, while I contemporaneously also worked on my habilitation. Through my continued contact with Ad Aertsen I completed the teaching parts at the University of Freiburg, where I eventually qualified as a professor  in neurobiology and biophysics in 2003. It took until 2006 before Markus and I managed to work and live in one place again.

Markus Diesmann: Now there is a piece of my life missing... For the build-up of the Bernstein Center, I had gone to Freiburg as a Juniorprofessor at the end of 2003. However, the university was, let’s say, hesitant to implement a tenure track. Also, the distance between Sonja in Berlin and me in Freiburg was not a perfect solution in the long run. The RIKEN Brain Science Institute in Wako City, Japan, then happened to offer new positions. Eventually we both applied, pointing out in the cover letters: we only come as a pair—but we want separate work groups...

Sonja Grün: ... which they wonderfully realized when we went there in 2006. We both had our own groups with separate rooms and budgets. We soon decided, however, to combine our spaces to share the infrastructure such as seminar room and computer cluster. To further deepen the links between our research areas of data analysis and modeling, we held joint lab meetings and journal clubs. At the beginning it was a bit bumpy, of course, as we both had our own styles of heading a group. But ultimately, we did a good job and it was an extremely good time.

Markus Diesmann: It was so productive and fun that we stayed until we got the offer from Jülich in 2011. The Forschungszentrum wanted to expand its neuroscience competences by a theory institute. Jülich is an ideal location. Besides the fact that we both have been offered a position, there are not so many places where you can combine neuroscience and supercomputing! Also the mission of the Helmholtz Association to create and maintain research infrastructure in long-term programs fits very well to our interests.

AG Diesmann und Grün

Group members of Markus Diesmann's and Sonja Grün's labs during a videoconference.
Copyright: Forschungszentrum Jülich

 

Now we have reached the present. What are your current research foci?

Markus Diesmann: One of my main interests is NEST, a simulation tool for spiking neuronal networks. Our goal is to simulate networks in natural size. A cubic millimeter of cortical tissue already comprises about 100,000 neurons and a billion synapses. This is a lot. However, if we take the largest supercomputer, we can today even simulate a billion neurons with NEST. Since the beginning, the basic principle of NEST has been that its technological development must be guided by neuroscience research. We only take the next developmental step if there is a concrete use case  or problem to solve.Today, NEST has become the leading simulation tool for large networks and is one of the simulation engines of the Human Brain Project.

Regarding neuroscience questions, we try to understand the layered structure of the cortex. How are structure and dynamics of a local circuit linked to larger brain structures? To a large extent, a local circuit is determined by synaptic input from other brain areas. We study networks that include both microscopic and macroscopic levels and analyze how brain areas are interconnected. This is where Sonja’s research comes into play. She examines data that have been recorded simultaneously in different brain areas using multi-electrode arrays.

Sonja Grün: You need to mention that the animal shows a complex behavior during the recording—it is the network interactions during natural behavior that are the most interesting part about it!

Markus Diesmann: Right. Towards the end of our time in Japan a new insight emerged from this. At that time, Sonja analyzed data from very complex studies and we realized that we were reaching a complexity barrier. Due to the numerous different steps, there are analysis chains that are hard to document and can run for weeks, putting the reproducibility at risk. Forschungszentrum Jülich wants to put the data analysis and reproducibility on a more firm footing. Sonja deals with the development of the analysis toolbox "Elephant" for the neurosciencific community and the more general problem of reproducibility.

Sonja Grün: Well, I would say, my research is roughly divided into three areas. First of all, I deal with the development of analysis methods for multi-channel data. In respect to statistics, it is not easy to discover correlations between activities in different brain areas or between neuronal activity and behavior. The second field are the close cooperations with experimental groups for which we do analyses. For instance, we collaborate with Alexa Riehle in Marseille and with Hiroshi Tamura in Osaka, Japan. And my third research pillar is the reproducibility of analyses. Over time, data analysis has reached such a degree of complexity that it is almost impossible to regenerate what others have done. In the simplest case this risk arises when a graduate student leaves the lab. We work on development of neuroinformatics tools that allow all members in a lab to evaluate the data with identical software. "Elephant" is an open source community project, which also allows scientists to integrate their own tools. Another aspect of this work is the gathering and representation of meta-data required to make the raw experimental data a meaning. One of my long-term goals is to establish a software that can analyze both experimental data as well as simulated data from Markus.

 

How would you describe your everyday research?

Markus Diesmann: Our research is based on communication. Here in the institute, we try to make arrangements to facilitate this. For example, there is an extra wide hallway, whiteboards are everywhere, and we have several videoconferencing systems. We always try to work together with many people on a common problem. One of our ideas is that in future we can only make progress if we learn to work in larger groups. Thus, the characteristic feature of our everyday work is to organize communication. Sonja, do you want to add something to this?

Sonja Grün: No, you have described it quite well.

Markus Diesmann: Programming on my own is definitely falling short at the moment.

Sonja Grün: This holds for me unfortunately as well.

 

What do you do in addition to research?

Markus Diesmann: Is there really a scientist who does not work all the time…?

Sonja Grün: I like yoga, which I have been doing now for a long time. And besides that, I enjoy my garden, which helps me to relax and clear my mind.

Markus Diesmann: I don’t do yoga—I anyway have a heart rate of 50 Hertz...