Skip to main content
Figure 12 | BMC Systems Biology

Figure 12

From: Modeling metabolic networks in C. glutamicum: a comparison of rate laws in combination with various parameter optimization strategies

Figure 12

Clustering of the best parameter values of all reversible models. A cluster analysis is carried out on all parameter values that have a fitness of less than 25. Each row of the above heatmaps corresponds to one parameter of the respective model and each column gives one parameter set that has been obtained in one of the optimization runs. The cluster algorithm swaps rows and columns to group similar parameters and parameter sets next to each other. This procedure leads to stripes from the left to the right, meaning that many parameters are often set to similar values in all optimization runs. If there were rectangular blocks within the heatmaps, this would mean that some parameters are correlated, thus showing a similar behavior, but because this is not the case, all parameters are distinct from each other, which can also be seen from the flat dendrograms at the side of each heatmap. Also, the dark-colored figures show that most parameters are set to low values because zero corresponds to black. The lighter the color the higher the value. For details, histograms showing the parameter distribution for the three models [see Additional file 2].

Back to article page