Multiple Linear Regression for Reconstruction of Gene Regulatory Networks in Solving Cascade Error Problems

Table 5

Characteristics of the datasets tested in the experiment and the AUROC results.

	Set 1	Set 2	Set 3
Number of cascade motif genes in the tested datasets	363	360	160
Number of random genes in the tested datasets; (not redundant with cascade motif genes)	397	533	255
Total number of tested genes	760	893	415
Total number of possible edges	576,840	796,556	171,810
Total number of correct prediction (CORRECT_PRED)	119	10	825
Total number of incorrect prediction	27,526	253	170,985
AUROC	0.511	0.502	0.662
	Average: 0.5584

Note:
() Cascade motif genes (italic text) are referring to the gene itself. This is not similar to cascade motif.
() Number of cascade motif genes in the tested datasets are obtained by comparing the cascade motifs genes with the genes in the datasets.
() Total number of possible edges = Total number of tested genes × (Total number of tested genes – 1).