FidlTrack: high-fidelity structure-aware single particle tracking resolves intracellular molecular motion in organelles sensing APP processing.

Single Particle Tracking (SPT) is a powerful technique for elucidating the dynamic behaviours of macromolecules within live cells. However, SPT's application to subcellular environments is hampered by the error-proneness of tracking at high particle velocities and densities and the lack of tools to assess trajectory reliability. Here, we introduce FidlTrack, a methodology that benchmarks and improves SPT fidelity. It contains three modules: a parameter optimiser that uses synthetic ground truth SPT data to determine the fidelity-maximising experimental and tracking settings; Structure-aware tracking, that exploits the information provided by organelle structures to constrain particle tracking algorithms; And a tracking quality evaluator that detects, quantifies and removes error-prone ambiguous track segments. Together these tools allow the rational design of SPT experiments, resolving the motion in tight and convoluted organelles, and provide up to 2-fold enrichment in accurate data. We showcase FidlTrack's utility for reliably tracking proteins in the cytosol, mitochondria and endoplasmic reticulum (ER). Further, we demonstrate its efficacy by analysing ER protein dynamics at exit sites, resolving BACE1 amyloidogenic cleavage of the amyloid precursor protein and characterising the spatiotemporal binding dynamics of an ER-targeted intrabody. FidlTrack is provided as a universal open-access platform that can be incorporated into any SPT pipeline.