Prediction of bypassed DNA lesions overestimation and validation results onrad14Δ background yeast cells at 75 J/m2 UV dose. The figure shows the experimental measurements on rad14δ background yeast cells irradiated at 75 J/m2 UV dose and the comparison with the corresponding simulation results. (A) Western blots showing a comparison between time-course measurements executed on WT cells (left part) and rad14δ background cells (right part), of mono-, di- and tri-ubiquitylated PCNA isoforms (top part, denoted by α- Ub) and of non modified PCNA (bottom part, denoted by α-His), sampled from 0 to 5 h after UV irradiation. As the aim of this experiment was not to carry out a precise quantification of the PCNA ubiquitylated isoforms, but only to verify the prediction of computational analysis, it was conducted with a single repetition. (B) Comparison between the values of experimental data (, ) and the mean dynamics (, ) of mono- and poly-ubiquitylated PCNA isoforms emerging from 100 independent stochastic simulations, where , represent the sum of mono-, di- and tri-ubiquitylated PCNA amounts obtained from experimental data and simulation outcomes, respectively. Stochastic simulations were executed starting from the same initial conditions (see Table 2 for molecular amounts and Table 3 for reaction constants) and with an estimated number of DNA lesions equal to 15018. Colored areas indicate the amplitude of stochastic fluctuations around the mean values , . Data are plotted by using the units representation (see Additional file 5). (C) Comparison between the ratio of experimental ( and , left bars) and simulated ( and , right bars) ubiquitylated PCNA isoforms at every sampled time point. Mean and standard deviation bars of both experimental and simulated ratios are plotted by using the normalized representation (see Additional file 5).
Amara et al. BMC Systems Biology 2013 7:24 doi:10.1186/1752-0509-7-24