Home Research Sections Section B
Personal tools

Section B

Network Level


On the network level, in research section B, we will ask: How does the collective dynamics of biological neural networks drive processes of adaptive coding and learning? Projects in section B will use innovative approaches developed recently to study network dynamics. On the experimental side, in vivo fluorescence imaging of neural activity, genetically encoded fluorescent probes, and the emerging ability to optically perturb the operation of intact networks are driving a revolutionary transition to optophysiology in the study of network dynamics. Projects B1 (Fiala/Witt/Timme), B2 (Schild/Wolf), and B5 (Fiala/Wörgötter) will use these approaches to examine key open problems in network function. B1 and B2 will use optophysiological methods to examine and model odour representations in invertebrate (drosophila) and vertebrate (Xenopus) olfactory systems. B1 and B5 build on excellent prior work using light sensitive ion channels expressed in genetically selected cell populations to perturb and probe network dynamics and plasticity mechanisms in vivo and in behaving animals. For instance, project B5, uses the cell type specific expression of channel rhodopsin 2 to manipulate dopamine and other neuromodulators in behaving flies to examine the neuronal substrate of reward-based learning. We thus expect this project to lead to a fundamental advancement in the understanding of reward-based learning. It appears particularly promising to conduct studies perturbing dynamic network states in conjunction with theoretical projects addressing networks on an advanced level as we do in projects B3 and B4 described below. All three projects B1, B2, and B5 are combining theoretical and experimental components and will be performed in collaboration between experimental (Fiala, Schild) and theoretical groups (Timme, Wolf, Wörgötter). The thematic coherence and basis for inter-project exchange in section B is further strengthened by the common focus on olfaction as the model sensory modality in B1, B2, and B5. These studies are complemented by two projects B3 (Timme) and B4 (Levina/Geisel) that extend the mathematical theory and computational modelling of dynamic network states and their modification by learning. B3 will study the impact of dendritic spikes and nonlinearities on spiking neural network (SNN) dynamics and learning. B4 will develop a theory of systems in which neuronal avalanche dynamics and associative memory function coexist and will ask whether generic properties of such systems can explain the phenomenology of memory consolidation and potentially of cognitive aging. These studies build on excellent prior work of the theoretical groups of Timme and Geisel and address question in the dynamics of networks that are of highly topical interest in computational neuroscience. In close contact with the integrated experimental-theoretical projects B1, B2 and B5 they will substantially deepen the research perspectives of the network dynamics research section.

Projects involved: