Fig. 4
Characteristics of risk score patterns based on m6A regulators. A The univariable Cox regression analysis for predicting the prognosis of m6A regulators. Hazard ratio > 1 represents the risk markers for survival and hazard ratio < 1 represents the protective markers for survival. B The most regularized and parsimonious risk score pattern was built by multivariable Cox regression using the least absolute shrinkage and selection operator (LASSO) Cox regression analysis The grey solid vertical lines represent the partial likelihood deviance ± standard error (SE). The black vertical lines are drawn at the optimum values by minimum criteria and 1‐SE criteria. C Kaplan–Meier curves for low- and high-risk score patient groups (Log-rank test). D, E Evaluation of the relationship of the risk score patterns with overall survival status. Red dots represent death and green dots represent alive. F Receiver operating characteristic (ROC) curve represents the predictive efficiency of the risk score patterns. G Heatmap shows the expression levels of the m6A regulators in low- and high-risk score patients with BCa. N status, M status, T status, tumor stage, tumor grade, gender, and age were used as patient annotations. Red represents the high expression of regulators and blue represents the low expression. H, I Cox regression analyses of the clinicopathological factors and risk score patterns in patients with BCa from TCGA. Univariable Cox regression analyses (H), Multivariable Cox regression analyses (I).*P < 0.05, **P < 0.01