Construction and verification of nomogram prediction model
Univariate Cox regression and multivariate Cox regression analysis were
performed to verify whether the model can be used as an independent
prognostic factor in TCGA-COAD and GSE40976. The ”survivalROC” software
package of the R language was used to analyze the independent prognosis
in the TCGA-COAD and GSE40976 data sets. Time-dependent ROC analysis is
performed on influencing factors, and the sensitivity and specificity of
survival prediction are analyzed by genetic marker risk score. Area
Under Curve (AUC) can be used as an indicator of prognostic accuracy. If
not specifically stated, P Value <0.05 was considered
statistically different for survival analysis. According to the results
of multivariate Cox regression analysis, the R language ”rms”, ”Hmisc”,
”lattice”, ”survival”, ”Formula”, ”ggplot2”, ”SparseM” software packages
are used to calculate and visualize the nomogram. Carry out the
calibration curve and the C-index analysis to verify the predictive
ability of the nomogram. DCA is used to evaluate the net rate of return
of the nomogram in clinical practice.