Fig. 10.

Performance statistics for the reverse-engineering algorithm. False positive rate and fraction (a) of true edges recovered (b, in %) for set of 1000 genes analyzed using a sensitivity analysis are presented. Results using two cutoff values, 0.01 and 0.05, using a $p_{\text{value}}$ relative to a chance that a connection appears using a random binomial distribution. Union represents a union of sets of results obtained from both experiments, CV represents the cutoff metric based upon co-efficient variation mentioned in the text, and Int represents intersection of results obtained from both experiments