opambos/DeepTRACE: DeepTRACE v1.0.0

This Zenodo archive corresponds to DeepTRACE version v1.0.0 released via GitHub as described in the published article. DeepTRACE is the software accompanying Pambos, O.J. et al. DeepTRACE: Flexible Machine Learning for Analysis and Discovery in Single Molecule Tracks. bioRxiv (2025). https://doi.org/10.1101/2025.05.15.654348 The GitHub repository hosts the actively maintained DeepTRACE codebase. DeepTRACE is a tool for analysing long single-molecule tracking data in living cells using machine learning. It is designed for experiments that capture multi-stage biological processes by learning patterns of molecular behaviour from motion, subcellular location, and photometric readouts. Users can annotate tracks, construct bespoke models, segment molecular behaviours, and perform quantitative analysis to assist biological interpretation of single-molecule tracking data. Models can be trained on datasets containing a few hundred tracks without specialised hardware. DeepTRACE accepts arbitrary numerical readouts as input, enabling native handling of emerging single-molecule photometric techniques (e.g. smFRET, anisotropy).