During the first half of this funding period, we successfully adapted the experimental method and related statistics analysis of the DCM-time machine system. This system enables tracking the past and present transcriptome of neural stem cells at the single cell level. We defined the optimal induction window and chase time to achieve the best signal-to-noise ratio, allowing us to distinguish real DCM methylation from background noise. A promising outcome was the ability to differentiate neural cell types based solely on their DCM methylome.
To track the interaction of neural stem cells with their neighbors, we leverage the exosomal trafficking between cells for barcode delivery. Two strategies are being tested for barcode transfer and donor-acceptor labeling. The first strategy involves protein cargo and the second RNA cargo. To benchmark both strategies, we needed to clone the different constructs and set up a double reporter mouse line to be used to distinguish between donor and acceptor cells. In parallel to the cloning and production of the different construct we are testing different delivery systems for in vitro and in vivo transduction of adult v-SVZ neural stem cells.
We have also advanced the development of a mathematical theory to describe models with structure on metric spaces (based on measure theory applied to partial differential equations; collaboration with the group of Piotr Gwiazda, IMPAN, Poland). In parallel, we are working on the development of a theoretical approach to a numerical method for structured population models on a metric space. We hitherto applied convergence of the Bayesian method for parameter estimation. Finally, we have built the first transcriptomically-structured population model and its well-posedness analysis.
To study NSC system dynamics in the mouse we have constructed nonlinear models to identify systemic feedbacks. We have conducted an analysis of the models in terms of their dynamics and stability and model simulations and comparison with experimental data on wild-type and perturbed neurogenesis for model identification (identification of non-linear feedbacks). This development involved the implementation of various parameter estimation methods (in collaboration with the group of E. Kostin, Numerical Optimisation, IWR, University of Heidelberg). To explore potential information contained in observed data fluctuations we have implemented stochastic models and are now implementing the first preliminary model to study NSC system dynamics in zebrafish
We are very excited about our latest development in implementing a computational framework for parameter estimation in models coupling different structures (collaboration of AMC with numerical and statistical analysts Robert Scheichl, Heidelberg University and Heikki Haario, Laperantha University, Finnland). This involves the development of a statistical approach for parameter estimation for PDE models with heterogeneity of outcome. As a first test we have implemented it for a reaction-diffusion system of chemical reactions and preparing the results for publication (to be submitted).
