Figure 1

Summary of theoretical analysis. A) Schematic illustration of the experimental procedure inducing priming effect. B) An abstract three-node model was chosen to represent that the stimulus can activate two parallel pathways (through x₁ and x₂) which converge to the monitored readout (x₃). A set of corresponding ordinary differential equations (ODEs) were constructed, and a Metropolis sampling algorithm was used to search for parameter sets giving the priming effect in the high-dimensional parameter space. C) Computational studies show that the parameter sets leading to priming naturally divide into two regions, corresponding to different priming mechanisms. $\underset{ss}{\Delta }{x}_{1,LD}$ ($\underset{ss}{\Delta }{x}_{2,LD}$): change of x₁ (x₂) level at the end of LD treatment period compared to those of untreated cells. D) The right region in C) can be further discriminated according to the sample abundance distribution (relative to the maximum within each group) of $\underset{\text{max}}{\Delta }{x}_{2,LD+HD}-\underset{\text{max}}{\Delta }{x}_{2,LD}$ (difference between the maximum level of x₂ during the HD treatment period after LD pretreatment and that of x₂ without LD pretreatment), suggesting overall three priming mechanisms. Panels A and B are adapted from [25].