Figure 1
From: Knowledge of the perturbation design is essential for accurate gene regulatory network inference
Workflow of the benchmarking pipeline. First, in silico true GRNs and perturbation-based gene expression data are generated, followed by addition of low, medium and high levels of Gaussian noise. GRNs are then inferred using P-based (that use the perturbation design) and non P-based (that do not use the perturbation design) inference methods, and finally the accuracy of each prediction is calculated by several measures, including area under precision-recall (AUPR).
