Runway: A novel, forward-screen method of analyzing catwalk gait analysis data

Abstract

Assessing locomotor endpoints is a crucial component of neurotrauma research. One method of measuring locomotor behavior is through gait analysis, where one empirically measures different aspects of animal locomotion. A common tool for performing this analysis is via using the CatWalk kinematic testing apparatus, which utilizes data captured from freely moving animals traversing a tunnel with a transparent walkway, specially lit and filmed from below to create a high-contrast video of paw placement during locomotion. The CatWalk software outputs high dimensional data, which presents an issue when attempting to determine what the sum of the data means in terms of locomotor outcomes. To address this concern, we have developed runway, an R package for the processing, analysis, visualization, and ultimately interpretation of Catwalk data. The runway pipeline uses a novel, forward-screen method for analysis, systematically identifying variables affected by the treatment or intervention in question and analyzes the isolated variables through various methods to pinpoint how animal gait is affected. We explain in detail how runway processes experimental data, and provide a pair of use cases to demonstrate runway’s power as an analytical tool.