Skip to main content
Fig. 12 | BMC Systems Biology

Fig. 12

From: Bayesian uncertainty analysis for complex systems biology models: emulation, global parameter searches and evaluation of gene functions

Fig. 12

The results from the history match of the k 2c =0 model, showing the waves 1, 3 and 5 runs as the purple, green and red lines respectively, for the 18 model outputs of interest. We see that a large number of acceptable wave 5 runs have been found that match the target data (the black error bars) and hence that the reduced k 2c =0 model is still consistent with the specified observational data, within a small region of the input space. Comparison with Fig. 10 shows the main differences between the two models, most noticeably in the Auxin_mufe output (7th error bar from the left) for which the vast majority of the input space returns an incorrect positive trend for the k 2c =0 model, as can be seen by the large number of wave 1 and 3 lines above zero, with only a small and hence hard to find region returning the correct negative trend. The k 2c >0 model conversely, always returns the correct negative trend

Back to article page