Skip to main content
Fig. 3 | BMC Systems Biology

Fig. 3

From: Inference of spatiotemporal effects on cellular state transitions from time-lapse microscopy

Fig. 3

Features regulating the transition rate λ can be estimated non-parametrically from cellular genealogies with annotated state transition events. a The transition rate estimated from 100 genealogies (posterior mean, black line) agrees well with the true constant transition rate (red line). Gray areas indicate the 95 % credibility region of the estimate. b The transition rate estimated from 100 genealogies simulated with linear time-dependent rate agrees well with the true rate (red solid line). c The transition rate as a function of local cell density ρ for 100 genealogies simulated with density-dependent rate. The estimated transition rate seems to depend on both local density ρ (in line with the simulated form λ=b·ρ) and time (see inset). d The estimated transition rate \(\hat \lambda \) as a function of both density and time reveals that the time-dependence observed in the inset in c is an indirect influence (density increases with time, see inset). Instead, the transition rate depends only on local cell density ρ (as seen by the predominantly uniform pattern of \(\hat \lambda \) in time for fixed ρ, indicated by arrow)

Back to article page