Fig. 1

a Calibration plots for the Taylor et al.¹³ and Wells et al.¹² models in the UK Biobank. The original models were initially calibrated to the UK Biobank population and following this the genetic risk score (GRS) was combined with the model’s original coefficient(s). To ensure comparable calibration between models with and without the GRS, we then further recalibrated by the predicted log hazard from the original model as a covariate in a flexible parametric survival model by itself, and with the addition of the GRS. b Change in the 5-year predicted probabilities (expressed as a percentage) of the recalibrated models after the addition of the genetic risk score. The x-axes are the predicted probabilities from the original models, and the y-axes are the difference in predicted probabilities between the GRS-augmented models and the original models. Histograms display the distribution of data along each axis. Note that the ranges of the axes differ between the two panels. The crowding of points close to the horizontal line at 0 on the y-axis illustrates that the addition of the GRS did not affect the predicted probabilities for the majority of participants