Algorithm performances. (a) Computational time of the algorithm compared with an estimation of the exhaustive search for different values of P on the E.coli core metabolism. (b) The influence of the size of the network on the performances is here reported as distribution of the computational time for the 6 metabolic networks listed in Table 2. Each distribution is built over 100 random samples obtained by changing the targets of a fixed number of drugs and keeping the precision parameter and the threshold constant (P = 2 and τ = 0.35). (c): Averaged computational time (same color code of the corresponding distributions reported in panel (b)) and error bar as a function of the size of the network (expressed as number of reactions, see Table 2). All tests have been carried out using the software ILOG-IBM CPLEX on a 2.3 GHz CPU.