Multidimensional phenotyping predicts lifespan and quantifies health in C. elegans

Abstract

Ageing affects a wide range of phenotypes at all scales, but an objective measure of ageing remains challenging, even in simple model organisms. We assumed that a wide range of phenotypes at the organismal scale rather than a limited number of biomarkers of ageing would best describe the ageing process. Hundreds of morphological, postural and behavioural features are extracted at once from high resolutions videos. A quantitative model using this multi-parametric dataset can predict the biological age and lifespan of individual C. elegans. We show that the quality of predictions on a held-out data set increases with the number of features added to the model, supporting our initial hypothesis. Despite the large diversity of ageing mechanisms, including stochastic insults, our results highlight a robust ageing trajectory, but variable ageing rates along that trajectory. We show that healthspan, which we defined as the range of abilities of the animals, is correlated to lifespan in wild-type worms.