Search for: All records

Abstract Extracellular matrices of living tissues exhibit viscoelastic properties, yet how these properties regulate chromatin and the epigenome remains unclear. Here, we show that viscoelastic substrates induce changes in nuclear architecture and epigenome, with more pronounced effects on softer surfaces. Fibroblasts on viscoelastic substrates display larger nuclei, lower chromatin compaction, and differential expression of distinct sets of genes related to the cytoskeleton and nuclear function, compared to those on elastic surfaces. Slow-relaxing viscoelastic substrates reduce lamin A/C expression and enhance nuclear remodeling. These structural changes are accompanied by a global increase in euchromatin marks and local increase in chromatin accessibility atcis-regulatory elements associated with neuronal and pluripotent genes. Consequently, viscoelastic substrates improve the reprogramming efficiency from fibroblasts into neurons and induced pluripotent stem cells. Collectively, our findings unravel the roles of matrix viscoelasticity in epigenetic regulation and cell reprogramming, with implications for designing smart materials for cell fate engineering. 

We consider a task of surveillance-evading path-planning in a continuous setting. An Evader strives to escape from a 2D domain while minimizing the risk of detection (and immediate capture). The probability of detection is path-dependent and determined by the spatially inhomogeneous surveillance intensity, which is fixed but a priori unknown and gradually learned in the multi-episodic setting. We introduce a Bayesian reinforcement learning algorithm that relies on a Gaussian Process regression (to model the surveillance intensity function based on the information from prior episodes), numerical methods for Hamilton-Jacobi PDEs (to plan the best continuous trajectories based on the current model), and Confidence Bounds (to balance the exploration vs exploitation). We use numerical experiments and regret metrics to highlight the significant advantages of our approach compared to traditional graph-based algorithms of reinforcement learning.