Self-modeling structure of evoked postsynaptic potentials.

Abstract

With the simplicity of the synaptic structure and physiology at neuromuscular junctions (NMJs) of crayfish and the given transmitter being released in quantal packets, a detailed assessment in the fundamental processes of chemical synaptic transmission is possible. Since the quantal event is the basic element of transmission, we consider an approach to further understand the characteristics of quantal responses. In this study, we introduce a method for combining information across excitatory postsynaptic potentials (EPSPs) that are quantal in nature. The method is called self-modeling regression, known in the statistics literature as SEMOR. This method illustrates that the differing timing and heights of EPSPs can be described with four coefficients measuring affine (shift and scale) transformations of the x and y axes. We demonstrate that this relationship allows us to provide a unified schema for the many functionals currently used in the literature, such as peak amplitude, tau, latency, area under the curve, or decay time. Computer code in R is available on the internet to perform the analysis.