Fig. 2: Ambiguity scoring for evaluating and improving tracking fidelity. | Nature Communications

Fig. 2: Ambiguity scoring for evaluating and improving tracking fidelity.

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

Fig. 2: Ambiguity scoring for evaluating and improving tracking fidelity.

a Percentage of linking errors occurring at ambiguous displacements in freespace simulations at the optimal linking distance for a wide range of characteristic lengths and spot densities. b Sketch depicting the ambiguity concept for a displacement: non-unambiguous displacements occur when only one spot is present in the next frame within the linking distance of the starting spot, leaving a single option for the tracking algorithm (top); When more than one spots are present in the next frame within the linking distance of the starting spot, the displacement is ambiguous with multiple linking options (bottom). c Map showing the Ambiguity Score (percentage of ambiguous displacements) in freespace simulations at the optimal linking distance for a wide range of characteristic lengths and spot densities. d Sketch presenting the trajectories splitting procedure for removing ambiguous displacements: each time an ambiguous displacement appears in a trajectory two new trajectories are formed containing all the displacements before and after the ambiguous displacement. e Percentage enrichment of correct displacements after ambiguity removal in freespace simulations at the optimal linking distance for a wide range of characteristic lengths and spot densities. f Non-ambiguous displacements density (mean ± SEM) as a function of the spots densities for different characteristic lengths (coloured) in freespace simulations. g Experimental single-particle trajectories data of an ER luminal probe HaloER stained with PA-JF646 dye from two successive recordings of the same cell at a “lower” (left) or “higher” (right) spots densities, colour-coded by individual trajectories. Spots appearing outside the structures (Fig. S3g) were removed. h Bar plot presenting the percentage of ambiguous displacements for the two recordings from (g). i Bar plot presenting the percentage of trajectories containing at least one ambiguous displacement for the two recordings from (g). j Spatial map presenting the percentage of erroneous displacements in each square bin (of size 50 nm) for the “low” (left) and “high” (right) density recordings from (g). k Bar plot presenting the recovered number of displacements for the two recordings from (g) before and after ambiguity removal. The results derived from simulations are averages over 5 repeats in freespace. Source data are provided as a Source Data file.

Back to article page