Glycosylation-Related Genes Predict the Prognosis and Immune Fraction of Ovarian Cancer Patients Based on Weighted Gene Coexpression Network Analysis (WGCNA) and Machine Learning

Table 1

Clinicopathologic information of training set and test set.

Covariates	Type	TCGA ()	Training ()	Test
Vital status	Alive	146 (38.93%)	80 (38.46%)	66 (39.52%)	0.834^a
	Dead	229 (61.07%)	128 (61.54%)	101 (60.48%)
Age	≤60	206 (54.93%)	105 (50.48%)	101 (60.48%)	0.053^a
	>60	169 (45.07%)	103 (49.52%)	66 (39.52%)
Grade	G1	1 (0.27%)	1(0.48%)	0 (0%)	0.577^b
	G2	42 (11.2%)	20 (9.62%)	22 (13.17%)
	G3	321 (85.6%)	181 (87.02%)	140 (83.83%)
	G4	1 (0.27%)	1 (0.48%)	0 (0%)
	Unknown	10 (2.67%)	5 (2.4%)	5 (2.99%)
Stage	Stage I	1 (0.27%)	1 (0.48%)	0 (0%)	0.451^b
	Stage II	22 (5.87%)	13 (6.25%)	9 (5.39%)
	Stage III	292 (77.87%)	158 (75.96%)	134 (80.24%)
	Stage IV	57 (15.2%)	33 (15.87%)	24 (14.37%)
	Unknown	3 (0.8%)	3 (1.44%)	0 (0%)

^aPearson’s chi-square test; ^bFisher’s exact test.