A computational framework for gene regulatory network inference that combines multiple methods and datasets

BMC Systems Biology

Table 2 Accuracy of GRN inference by integrating gene KO datasets in the MOO framework

Type	Size	Ratio methods MOO-Sr/MOO-Tr	Corr. methods MOO-Sc/MOO-Tc	Ensemble MOO S_ens/MOO T_ens
Undirected	20	0.77/0.70	0.70/0.65	0.77/0.70
Undirected	35	0.88/0.79	0.75/0.70	0.88/0.79
Undirected	50	0.85/0.79	0.75/0.69	0.85/0.79
Directed-signed	20	0.23/0.24	0.42/0.36	0.47/0.39
Directed-signed	35	0.18/0.21	0.54/0.49	0.69/0.57
Directed-signed	50	0.17/0.22	0.54/0.45	0.64/0.51

The table shows the AUC values obtained for 20, 35 and 50-gene networks, for undirected and direct-signed networks with MOO procedures integrating time course and gene KO datasets. Ratio, correlation and ensemble-based methods are shown in separate columns. AUC values for different procedures within each column are separated by a forward slash. Note that we marked the AUC values in bold to highlight the opposite trend in inference accuracy of the ratio and correlation procedures.

ISSN: 1752-0509