A pseudo-value regression approach for differential network analysis of co-expression data

BMC Bioinformatics

Table 5 Scenario III simulation results of binary group variable in the multivariable and univariable robust regression model using pseudo-value approach with 1000 replicates

p	n	Precision				Recall				F1
p	n	PRANA (Mult)	PRANA (Univ)	dnapath	DINGO	PRANA (Mult)	PRANA (Univ)	dnapath	DINGO	PRANA (Mult)	PRANA (Univ)	dnapath	DINGO
20	40	0.67	0.60	0.64	0.58	0.57	0.74	0.50	0.75	0.59	0.65	0.55	0.64
	100	0.67	0.58	0.61	0.58	0.65	0.85	0.73	0.79	0.64	0.68	0.65	0.66
	200	0.66	0.58	0.59	0.58	0.76	0.91	0.85	0.80	0.69	0.70	0.69	0.67
	500	0.64	0.57	0.58	–	0.87	0.95	0.95	–	0.73	0.71	0.71	–
	1000	0.64	0.57	0.58	–	0.92	0.97	0.97	–	0.75	0.71	0.72	–
50	40	0.57	0.50	0.54	0.49	0.47	0.62	0.33	0.67	0.50	0.54	0.40	0.55
	100	0.58	0.49	0.52	0.49	0.50	0.71	0.51	0.79	0.52	0.58	0.51	0.60
	200	0.58	0.49	0.51	0.49	0.52	0.76	0.60	0.83	0.53	0.59	0.54	0.61
	500	0.55	0.48	0.48	–	0.67	0.88	0.83	–	0.59	0.61	0.60	–
	1000	0.54	0.48	0.48	–	0.81	0.93	0.91	–	0.64	0.62	0.62	-
100	40	0.57	0.49	0.53	0.48	0.44	0.57	0.22	0.75	0.49	0.52	0.30	0.58
	100	0.58	0.49	0.53	0.48	0.47	0.64	0.35	0.76	0.51	0.55	0.41	0.58
	200	0.58	0.49	0.53	0.48	0.46	0.65	0.40	0.80	0.50	0.55	0.45	0.60
	500	0.56	0.47	0.49	–	0.44	0.74	0.60	–	0.48	0.57	0.53	–
	1000	0.53	0.47	0.47	–	0.63	0.85	0.82	–	0.57	0.60	0.59	–

The network structure depends on age covariate by unequal sampling proportion depending on age categories. Random network is generated at each simulation replicate. Sample size \(n = (500, 1000)\) or gene size \(p = 100\) were not included for DINGO due to heavy computational costs. The best results are highlighted in boldface

ISSN: 1471-2105