AutoGaitA: A versatile quantitative framework for kinematic analyses across species, perturbations and behaviours

Individual behaviours require the nervous system to execute specialised motor programs, each characterised by unique patterns of coordinated movements across body parts. Deep learning approaches for body-posture tracking have facilitated the analysis of such motor programs. However, translating the resulting time-stamped coordinate datasets into meaningful kinematic representations of motor programs remains a long-standing challenge. We developed the versatile quantitative framework AutoGaitA (Automated Gait Analysis), a Python toolbox that enables comparisons of motor programs at multiple levels of granularity and across tracking methods, species and behaviours. AutoGaitA allowed us to demonstrate that flies, mice, and humans, despite divergent biomechanics, converge on the age-dependent loss of propulsive strength, and that, in mice, locomotor programs adapt as an integrated function of both age and task difficulty. AutoGaitA represents a truly universal framework for robust analyses of motor programs and changes thereof in health and disease, and across species and behaviours.