A posture-based Markov analysis of behavioral states in Caenorhabditis elegans

Abstract

Organisms of varying degrees of complexity coordinate their diverse behavioral outputs over time, yet the internal neural dynamics underlying such behavioral organization is not completely understood. Behavior coordination can be captured through the quantitative description and subsequent analysis of behavioral states. Here, we develop an analytical method for the characterization of behavioral states in C. elegans. We observe posture sequences in wild type C. elegans and utilize a hidden Markov model to detect the behavioral states giving rise to these posture sequences. We then demonstrate that this method is generalizable to different C. elegans strains by applying this posture-based Markov analysis to C. elegans mutants and survey how these mutants differentially exhibit key behaviors within these behavioral states. This methodology provides a framework by which behavioral states can be quantified for further study of the neural dynamics underlying behavior coordination.