Table 2.

Gaussian mixture model fits to instantaneous velocity distributions

	Experiments	Basic model	Best-fit model
Central peak, nm/s (%)	−18 (65)	127 (53)	−8 (93)
Plus-end peak, nm/s (%)	98 (23)	587 (6)	452 (0.5)
Minus-end peak, nm/s (%)	−204 (11)	−334 (41)	−319 (6.5)
Residual sum of squares (RSS)	–	3.93 × 10−5	2.76 × 10−6

Experimental data from Feng et al. (34) include N = 30 traces, and basic and best-fit models both include N = 1000 traces. Percentages are the relative integrated weights of each mode. Fits used the “fitgmdist” function in MATLAB with three Gaussians having initial values of 0, 500, and −300 nm/s. RSS values denote the residual sum of squares between the probability density functions each model and the experimental data.