Application of artificial intelligence strategies to the analysis of neurotransmitter receptor dynamics in living cells

Abstract

Abstract: Storm (stochastical optical reconstruction microscopy), a form of single-molecule nanoscopy, calls for a variety of statistical and mathematical operations to reconstruct the original objects from their noisy wide-field point spread functions [1]. We are interested in understanding the dynamics of the nicotinic acetylcholine receptor (nAChR) protein, a cell-surface neurotransmitter receptor. Analyzing the translational motion of nAChR molecules by single-particle tracking in living cells is a complex task. In order to understand how nAChR molecules associate/dissociate into/from nanometer-sized clusters over time, and to characterize their trajectories according to different mathematical models, we are developing analytical procedures based on artificial intelligence. Due to their speed of calculation and accuracy, deep learning models are clearly an improvement on classical models in biological image analysis and biomedical science.