Fig. 4
From: The phenotype control kernel of a biomolecular regulatory network
Comparison of target control methods. a Here, the cancer cell signaling network denotes the simplified MAPK network. The desired phenotype value is Proliferation = 0. We named the target control method for finding driver nodes as ‘Structural Controllability’ in [9] and the control method for finding all minimal control sets as ‘Converging Tree’. In the top box, red and orange balls represent driver nodes and Proliferation nodes, respectively, where the dotted red arrows denote links contained in the maximum matching. The number ‘1’ before the green bar denotes the number of driver nodes. In the bottom box, purple circles, yellow and orange balls represent layered nodes, two input nodes and Proliferation nodes, respectively, where each dotted red arrow from a layered node in the ith layer denotes a link directed to other layered node in the ith or a higher layer. The numbers ‘1’ and ‘2’ before the blue bars denote the numbers of control nodes in a control set, which can be found from the converging tree. b The numbers ‘1’, ‘3 and 8’ next to the green and blue bars denote the numbers of possible control sets obtained from the two target control methods, respectively. The converging tree consists of 11 control sets up to the last level, where each of 3 control sets {(JUN, FOS)= (1,1)}, {(JUN, ATF2)= (1,1)} and {(JUN, p38)= (1,1)} has two control nodes, and the other 8 control sets are singleton sets