Figure 6From: Reconstruction of large-scale regulatory networks based on perturbation graphs and transitive reduction: improved methods and their evaluationPerformance of the novel inference techniques on the S. cerevisiae dataset validated against four silver standards. The plots show the AUPR and the number of true positive edges computed for the 500 best-ranked edges against four “silver standard” networks (see text for explanations). Parameters used to infer the networks are β = 2.0 and γ = 0.05 for the PG, α = 0.95 for TRANSWESD and α = 0.15 for LTR.Back to article page