Fig. 2

Estimation of the start time of chloroplast movement using three approaches. A Bayesian inference of the state-space model for a chloroplast in cell 1. The observed distance of the chloroplast from the blue microbeam (first panel), the observed and inferred velocity of movement (second panel), the inferred coefficient of the blue microbeam (third panel), and the inferred random fluctuations of the velocity (last panel) are shown. B Estimation of the state-space model for a chloroplast in cell 1 using the Kalman filter. The observed distance of the chloroplast from the blue microbeam (first panel), the observed and inferred velocity of movement (second panel), and the inferred coefficient of the blue microbeam (third panel) are shown. C Estimation of the start time for a chloroplast in cell 1 without the state-space model. The observed distance of the chloroplast from the blue microbeam is shown. D Comparison of the start time of chloroplast movement between the visual estimation and the three methods. Each dot represents the data point of each chloroplast. E Comparison of the total time required to estimate the start time of the movement of 11 chloroplasts in cell 2 between the visual estimation and the three methods. In the panels including inferred values, dots, solid lines, and shaded regions are the observed values, medians, and 99% credible intervals of the Bayesian inference, respectively. In A–C, orange solid lines and green dashed lines represent the start time estimated by each method and the visual observation, respectively. The shaded and light regions are the periods under observation light and under blue microbeam irradiation, respectively