Discovering the Structure of Odorants in Caenorhabditis Elegans using Formal Concept Analysis

Abstract

Embodied and Embedded cognition posits that animals act by operating a sensory-motor loop over an environment through the concourse of their nervous systems. A notable challenge in studying the nervous system lies in the absence of a model of its sensorium, the mechanism whereby an organism perceives its surroundings. This paper addresses the inquiry into the perception mechanisms of Caenorhabditis elegans as a model organism. To this end, we have delimited the perceptual domain to 23 distinct odorants, for which we have the corresponding neural activity data from the 11 pairs of sensory neurons. Thispaper elucidates how the integration of various FCA-related methodologies—specifically, K-FCA max-plus and min-plus analyses, as well as cosine distance calculations—allows for the obtention of a representation of the perceptual framework of C. elegans. Additionally, we propose avenues for further refinement, including the consideration of the idempotent singular value decomposition within the FCA framework to improve said perceptual representations.