"Synapse” (2006) No. 60 p. 32-44

In “The self-modeling structure of evoked post-synaptic potentials,” Kert Viele and other authors discuss synaptic transmission and how it can be studied by analyzing the electrical signals produced through stimulation of the neuromuscular junction. The article illustrates how self-modeling regression (SEMOR) can be used to unify previous analyses of such data. The central idea is to assume a common shape function for the entire series of electrical signals, then align all of the electrical signals to this common shape.

The resulting analysis accounts for 98 percent of the variation in the data. In addition, the self-modeling structure allows the 200 functionals produced by standard software to be described by four self-modeling coefficients, which improves the interpretability of the analysis. As a final benefit, the self-modeling structure provides improved estimates of the underlying shape function.

Kert Viele is a professor of statistics in the College of Arts and Sciences at the University of Kentucky. He received his bachelor’s, master’s and doctorate degrees in statistics from Carnegie Mellon University in Pittsburgh, Pa. In 2001, Viele won the UK Provost's Award for outstanding teaching. His expertise includes the interfaces between model selection, mixture modeling and functional data analysis — usually in the context of proteomic or neurological data.