Fig. 1

LASSO and Cox regression models. a LASSO Cox model fitting. Each curve represents a gene. The profiles of coefficients were plotted versus log(λ). Vertical lines indicate the positions of 7 genes with coefficients > 0, as determined by tenfold cross-validation (n = 480). b λ was determined from tenfold cross-validation. The x-axis represents log(λ); the y-axis represents binomial deviance. Optimal values calculated from minimum criteria and 1 SE of the criteria are indicated by the dotted vertical lines. c, d Univariate forest plot (c) and multivariate forest plot (d) showing association between COX expression and clinical features linked to OS in CRC (n = 477). e, f Univariate forest plot (e) and multivariate forest plot (F) showing the association between COX expression and clinical features related to DSS in CRC (n = 461). g, h Univariate forest plot (g) and multivariate forest plot (h) showing the association between COX expression and clinical features related to PFI in CRC (n = 477). CI Confidence interval, HR hazard ratio, SE standard error; for other abbreviations, see Abbreviation list