Simulation framework on developing neural network model of C. Elegans

Abstract

The nervous system is a complex yet efficient structure – with superior information processing capabilities that surely surpass any man-made high-performance computer. However, without having a complete architectural blueprint of this “technology,” understanding its underlying mechanisms is a challenging task. Current research has focused on investigating the interactions between neurons. However, this is not sufficient to explain each neuron’s functionality – as they might be physiologically different. One of the approaches to understanding the role of a specific neuron is to observe how its behaviour changes as a network develops. Notwithstanding, observations of biological changes in behaviour accompanying network developmental are experimentally challenging. This therefore creates a new possibility for research exploration: using a network developmental model that can simulate the behaviour of the biological network during the development process. In this research, the biological network of C. Elegans is used as the foundation for the design of the model. Although composed of only with 302 neurons, the nematode's network has the capability of handling complex biological processes, for instance, locomotory and sensory behaviour. In addition, the invariant developmental cell lineage and available network connectivity information provide advantages in terms of analysing the developmental pattern. This creates the possibility of defining a mathematical description of the developmental trajectory. This work aim to design a network trajectory model that is governed by specific rules – a model that has the ability to self-generate a single cell for the multicellular network. For this purpose, a high-processing event-based neural simulator with a 2D virtual environment is developed – to be used for the simulation of a complex multi-stimulus environment. This paper describes the approach and initial design of the network developmental trajectory model.