Fig. 5

Example of network optimization: best run. a Evolution of the fitness function F = (f _T , − N _ess. nodes , N _nodes , − N _edges ) during one run of the optimization method using the training set 1 given in Additional file 2, the PKN given in Additional file 4 (PKN 1) and the 14 essential nodes from the reduced cell-fate decision model proposed by Calzone and co-authors [11]. For each given iteration, blue dots correspond to individual replicas and black dot to the replica with best value of the fitness function. b Network with best value of the fitness function obtained during this run. The run shown in this figure is the one that produced the network with minimal value of the fitness function among 500 independent runs