A neural network model enables worm tracking in challenging conditions and increases signal-to-noise ratio in phenotypic screens

High-resolution posture tracking of C. elegans has applications in genetics, neuroscience, and drug screening. While classic methods can reliably track isolated worms on uniform backgrounds, they fail when worms overlap, coil, or move in complex environments. Model-based tracking and deep learning approaches have addressed these issues to an extent, but there is still significant room for improvement in tracking crawling worms. Here we train a version of the DeepTangle algorithm developed for swimming worms using a combination of data derived from Tierpsy tracker and hand-annotated data for more difficult cases. DeepTangleCrawl (DTC) outperforms existing methods, reducing failure rates and producing more continuous, gap-free worm trajectories that are less likely to be interrupted by collisions between worms or self-intersecting postures (coils). We show that DTC enables the analysis of previously inaccessible behaviours and increases the signal-to-noise ratio in phenotypic screens, even for data that was specifically collected to be compatible with legacy trackers including low worm density and thin bacterial lawns. DTC broadens the applicability of high-throughput worm imaging to more complex behaviours that involve worm-worm interactions and more naturalistic environments including thicker bacterial lawns.Author summary: Measuring how animals move in videos is useful in genetics and neuroscience experiments. Humans are good at following moving animals, but computers struggled until around ten years ago when deep neural networks provided a way to recognise objects, including animals, if they were trained on many examples. Ironically, these methods have not always worked as well for the simplest animals like nematode worms because they don’t have clear keypoints on their bodies like joints that the networks can learn to recognise. In this paper we show that a recently developed network that was designed specifically for slender objects like worms can be trained to recognise and track worms crawling on agar plates (a common lab environment) even when they are in thick food or overlapping with each other. This network uses information from a series of frames, instead of single images, to resolve difficult cases. Better tracking makes it easier to detect difference between worms treated with different chemicals which will improve future drug screens.