CNCR aims at understanding the healthy and diseased brain at various description levels, ranging ‘from molecule to mind’. For this, a systems biology approach is prerequisite.
Our Systems Biology approach is focused to the analysis of the synapse in order to provide functional understanding from protein interactions to synapse physiology and from neuronal network properties to behavior.
High-throughput experiments
CNCR uses state-of-the-art techniques that allow large screens of mouse mutants and parallel assessment of parameters at various levels of function in the brain. These include protein analyses of the synapse (multi-dimensional liquid chromatography, iTRAQ (quantitative) labeling, and high-throughput mass spectrometry), subsynaptic protein complex analysis (via immunoprecipitations and pull-down of tagged proteins), biacore interaction analysis to derive kinetic parameters for proteins of interest, synaptic localization of proteins using confocal and electron microscopy, connectivity analysis through hexa-patch recordings, and fully automated parallel behavioral analysis (automated home-cage analysis in collaboration with Sylics). In addition we use model systems like the autapse to standardize assays for studying the role of synaptic genes in neurotransmission in a European Systems Biology initiative (www.synsys.eu)
Databasing
Databases are used to store and manage large datasets generated in the CNCR, ranging from genetic, morphological, physiological and behavioral data. We are partner in two European initiatives (www.synsys.eu and www.eurospin.mpg.de) to integrate this data in central databases in Europe. In addition we have access to large databases with genomic data from patient cohorts world wide for genome wide association studies (GWAS).
Computational modeling
Within the CNCR computational modeling is applied to a broad spectrum of projects, including single compartment models for synaptic processes, interaction networks for synaptic proteins, and multi-compartmental models for neurons and networks. In some cases probabilistic methods like Bayesian models or mixture of probabilistic PCA’s are used to account for sparseness of parameter information and experimental variation in datasets. In all cases the aim is to integrate modeling and experimental research by using models to simulate experimental observations and predict new results. Models are implemented in C++, Python, Matlab or the simulation environment NEURON.
