Mathematical and computational models of the rat behavior into the Elevated Plus Maze

Author(s)

Hector Julian Tejada Herrera

• Document type: Doctoral Thesis
• Press: Ribeirão Preto.
• Institution: Universidade de São Paulo (USP). Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto (PCARP/BC)
• Defense date: 2010-05-28

Examining board members

Antonio Carlos Roque da Silva Filho; Leandro José Bertoglio; Ricardo Luiz Nunes de Souza; Renato Tinós

• Advisor: Antonio Carlos Roque da Silva Filho

Abstract

The Elevated Plus Maze (EPM) is an animal model for the study of anxiety, its biological foundations, and the effects of different kinds of drugs on behavior. A group of different modeling methods have been used to study the rat behavior in the EPM. One of them characterizes rat behavior the EPM using directed graphs and proposes an index which can be used to classify the drug type and dosis. The other two methods were used to construct computational models for the rat behavior. The first of these models used Markov chains to reproduce the rat behavior. This approach offers a method to characterize the rat behavior, which is able to identify the effects of certain kinds of drugs on the way the rat explores the maze and, at the same time, raises some clues on how much information is used by the animal to take a decision. The second one was built based on the idea that conflict determines the rat behavior in the EPM. Conflict was introduced in the model via an adapted version of Grossberg\'s gated dipole artificial neural network model. The goal was to evaluate the viability of a competition-based model with three kind of drives: a drive to explore threatening places, a drive to seek protection, and a drive to move, related to the energy of the animal. The model receives and processes signals from the environment according to the states of these three drives, and the way in which the signals are processed can be influenced by effects of specific drugs. This model reproduces the expected effects of three types of drugs with modifications of up to two parameters. The model also reproduces part of the expected behavior of the animal in the Arena, a different maze, by requiring only small adjustments to reproduce the trajectory of the rat around the walls. It is possible to conclude that this computational model captures elements of the interaction of variables which control the rat behavior in the EPM and how drugs interact with these variables. These elements may also be present in the rat behavior in the Arena. The tools and models presented here offer new paradigms to study rat behavior in the EPM. They can offer new benchmarks to characterize rat behavior and can be used to study the effects of different drugs and their interactions. (AU)