Affiliation - University of California Irvine
Title of Talk - Statistical learning of biophysical factors controlling signaling molecule localization in primary cilium
Many signaling cascades involve dynamic relocalization of signaling molecules at the primary cilium. Mislocalization of signaling molecules is associated with a class of diseases called ciliopathies. Numerous studies have been focusing on identifying different molecular factors which affect the distribution of signaling molecules in the primary cilium; however, even for the most well studied pathway, such as Hedgehog signaling pathway, the mechanism which selectively controls the localization of signaling molecules is still debated. Furthermore, there is much prior evidence for a diffusive barrier at the base of the cilium, but 'diffusive barrier' can mean several distinct biophysical phenomena, e.g., a mechanical barrier or an increase in local viscosity. Here we propose a primary cilium signaling molecule transport model with consideration of different molecular and biophysical factors which are hypothesized to be important for its coordinated transport and selectivity. With this model, we predict the distribution of signaling molecules upon various perturbations of biophysical factors. We further develop a method which uses single particle tracks to distinguish local changes in viscosity versus local elastic barriers. Moreover, our method can distinguish how much of the movement is due to membrane heterogeneity versus cytoplasmic (or cilioplasmic) structures. This method is based on advances in Bayesian statistical learning to detect subtle differences between biophysical forces which are difficult to be experimentally identified. We demonstrate our methods using synthetic data.