Effective modeling of a partial neural network of C. elegans: Theory and application

This paper presents an effective formulation for a partial neural network of Caenorhabditis elegans. Neurons in the whole neural network are divided into two complementary groups. One is the neurons in a partial network and the other is the remaining neurons. To evaluate the influence of the latter on the former, I consider the means and the variances of the remaining neurons instead of each dynamics of them. Equations for the means and the variances are derived by a dynamical mean-field approximation. Furthermore, the presented formulation is applied to predict the synaptic signs in the touch sensitivity network of C. elegans.