Fig. 1: Schema of the MCCP for the PRS-to-trait prediction. | Nature Communications

Fig. 1: Schema of the MCCP for the PRS-to-trait prediction.

From: Translating polygenic risk scores for clinical use by estimating the confidence bounds of risk prediction

Fig. 1

Using summary statistics from a reference GWAS, PRS is obtained for the target population. The calibration set includes individuals with known disease status. A model is built on the training set and nonconformity measures (NCM) can be computed for both calibration and test sets, respectively. The NCMs distributions in the calibration set are obtained for case and control, respectively (blue and red curves). The MCCP probability values for an individual to be the case (p1) or control (p0) (gray vertical bar) are estimated, respectively. In principle, population structures, age, sex, and other covariates can also be added to the model to increase its performance.

Back to article page