Fig 2: The risk model has the potential to predict metastasis in patients with OS. (A) Non-metastasis and metastasis cases between low- and high-risk group OS tissues in the TARGET and GSE21257 cohorts. (B) ROC analysis for the diagnostic value of the risk model in the prediction of OS tissue metastasis. (C, D) IHC was performed to detect the expression of COL22A1, CDK6, RNASE6 and AOC3 in non-metastasis and metastasis OS tissues (magnification 200× and 400×). *p < 0.05; **p < 0.01; ns, no significant.

Fig 3: The risk model exhibits high prognostic value in the GSE21257 verification cohort. (A) OS tissues in GSE21257 were divided into low- and high-risk groups according to median risk scores. (B) Kaplan survival analysis indicated the difference between low- and high-risk group OS tissues in GSE21257. (C–E) ROC analysis of the risk model for the 1-year, 3-year, and 5-year survival rates for OS patients in GSE21257. (F) Alive and death cases between low- and high-risk group OS tissues in GSE21257. (G) Expression of COL22A1, CDK6, RNASE6 and AOC3 between low- and high-risk group OS tissues in GSE21257.

Fig 4: Construction of a risk model. (A) Univariate COX regression analysis for the 94 genes associated with the infiltration of resting dendritic cells. (B, C) LASSO analysis for other important genes associated with the survival rate of OS. (D) The HR and p value of genes (including AOC3, CDK6, COL22A1 and RNASE6) under multivariate COX analysis are shown.