Fig. 5

Estimation of parameters in the Brownian ratchet scanning model by MCMC algorithms. A A flowchart illustrating decisions involving PIC movement, translation initiation (or leakage), placement of a “pawl”, and activation of the NMD pathway. While it was reported that PICs loaded on the mRNA could move further upstream (5′) to scan AUGs having extremely short UTRs, this scenario has not been considered in our simulation because we have not yet known all the features regarding the steric hindrance between eIF4E-cap and the PIC mRNA-binding channel. B Two PIC trajectories exemplify the simulated PIC scanning process along the mRNA. According to Archer et al. [47], the 13th–15th nucleotides of a PIC-binding mRNA fragment are inspected for complementarity to the Met-tRNAi anticodon. Therefore, the 13th nucleotide of a PIC-binding mRNA fragment is used to plot the PIC position. C Trace plots show changes in values of p.Pawl, p.Leakage, and p.NMD in an MCMC chain. Green dots mark the MCMC iterations that accepted a new parameter value due to a reduction in the RSS. D The observed GFP intensities of out-of-frame dATG variants in the yeast experiments (two replicates combined) and the simulated GFP intensities using one of the 30 sets of optimized parameters by the MCMC algorithms. The results of all 30 sets of optimized parameters are shown in Additional file 1: Fig. S12. E Two-dimensional density plot shows the distribution of the outcome values of p.Pawl and p.Leakage among the 30 MCMC chains (shown by dots). Two additional density plots involving p.NMD are shown in Additional file 1: Fig. S11C. F, G Estimated parameters and standard error (SE) in the Brownian ratchet scanning model