Projects of the second funding period (2010-2017)
Research in the second funding period was divided in two branches (A and B).
Branch A: Discharge patterning and variability in cellular neuroscience
The cellular branch (A) focuses on discharge patterning and variability. A set of closely related projects (Schmitz & Kempter; Brecht & Kempter; Schimansky-Geier, Heinemann, Gloveli & Schreiber) investigates discharge patterning and variability in the rodent hippocampus, which is an emerging key area of expertise of the BCCN Berlin. The successor of Prof. Herz will also contribute to this key field of research. In addition a set of projects aims at understanding information transmission and variability in small neural networks: Sigrist & Blüthgen investigate synaptic dynamics at different timescales by combined molecular, optic and computational modeling; Herzel & Kramer use computational tools to understand how sensory cues (in their case odors) entrain the mammalian circadian clock and, specifically, how precise timing can be derived from sloppy/noisy rhythms; Schreiber & Ronacher analyze information transmission and discharge variability in unmyelinated axons; Lewin, Schmuker & Brecht aim at understanding how consistent and subjectively different perception can be transmitted through overlapping mechanosensory afferents, thus constituting the computational problem of unmixing sensory channels; Nawrot & Pflüger investigate humoral modulation of neural networks in the insect nervous system.
A1 – Burst-Timing Dependent Plasticity (Schmitz, Kempter)
A2 – Spikelet Activity and Hippocampal Spatial Representations (Brecht, Kempter)
A3 – Dynamical Switching Between Network States in Hippocampal Area CA3 (L. Schimansky-Geier, T. Gloveli, U. Heinemann, S. Schreiber)
A4 – Theoretical Analysis of Hippocampal Memory Formation (Kempter)
A5 – Timescales of Synaptic Plasticity as a Basis for Learning and Memory (Sigrist, Blüthgen)
A6 – Precise Timing from Sloppy Rhythms – The Circadian Clock of the Olfactory Bulbs (Herzel, Kramer)
A7 – Unmixing of sensory channels encoding noxious mechanical and heat (Lewin, Schmuker, Brecht)
A8 – Coding Strategies in Unmyelinated Axons: Lessons from the Grasshopper Auditory System (Schreiber, Ronacher)
A9 – Computational Analysis of Modulatory Function in Small Motor Networks in Insects (Nawrot, Pflüger)
A10 - Neural Information Filtering (Lindner, Schreiber)
A11 - The Mathematical Analysis of Interacting Stochastic oscillators (Stannat)
A12 - Numerical Analysis and Simulation of Cooperative Phenomena in Interacting Stochstic Oscillators (Stannat)
A13 - Dynamics of inhomogeneous neural systems with nonlocal coupling (Hövel)
Branch B: Prediction in human neuroscience and cognition
Within branch (B) we pursue computational approaches to prediction in human neuroscience and cognition. Curio & Müller and Dahlem, Dreier & Schöll aim at understanding patterning and variability of large-scale brain signals in a biomedical context: Curio & Müller seek to predict behavioral variability from interareal interactions of non-invasively obtained macroscopic oscillatory EEG signals, integrating microcircuit concepts from branch-A projects of Schmitz / Kempter and Schimansky-Geier et al.; Dahlem, Dreier & Schöll will predict the spread of slow signals, e.g. in migraine, using non-linear modeling approaches. Another set of groups working on human neuroscience addresses the issue of prediction from a computational perspective. Wichmann & Engbert use machine-learning techniques to identify which stimulus features predict saccade targeting. This study will also touch the classic topic of perceptual stability in the face of rapidly-varying sensory signals, with a focus on noise that results from eye movements. Predictions in the cognitive domain are the research of several projects. Haynes, Brandt & Obermayer use a combination of psychophysics, computational neuroimaging and modeling to investigate how the brain extracts expectations from the statistics of sensory stimuli. As a vital factor the dependency of such expectations on reward will also be addressed. Similarly, a study by Heinz & Obermayer uses a probabilistic reversal learning task combined with a network model to investigate the role of reward prediction for clinical cases of motivational dysfunction and impulsivity. Finally, a study by Kurths & Blankenburg uses a combination of neuroimaging and modeling to investigate whether predictive coding provides an explanation for how sensory signals are constrained by contextual a priori knowledge, with this top-down approach in the somatosensory system efficiently converging with the bottom-up analysis of Curio & Müller as well as the predictive bottom-up approach of Wichmann & Engbert.
B1 – Spatiotemporal Modeling of Macroscopic EEG Rhythms (Curio, Müller)
B2 – Spreading Depolarizations in Stroke, Migraine, and Epilepsy: Theory and Experiment (Dahlem , Dreier, Schöll)
B3 – Perceptual Stability in the Face of Input Variability: Inferring Critical Features from Human Saccade Targets (Wichmann, Engbert)
B4 – The interaction of expectation and value in human visual cortex: Psychophysics, neuroimaging and computational modeling (Haynes, Obermayer, Brandt, Martin)
B5 – Reward-Motivated Learning and Memory across the Lifespan (Heinz, Obermayer)
B6 – Testing Predictive Coding as a Principle Mechanism for Perceptual Inference and Learning (Blankenburg, Kiebel, Kurths)
B7 - Large-scale neuronal models for functional networks of the human cortex (Hövel)
Charité Universitätsmedizin Berlin, Freie Universität Berlin, Humboldt-Universität zu Berlin, Technische Universität Berlin, Max-Delbrück-Centrum für Molekulare Medizin Berlin-Buch, Universität Potsdam.