Fig. 4
From: Interplay of Mendelian and polygenic risk factors in Arab breast cancer patients
Evaluating the biological plausibility of the calculated PRS. a The Spearman correlation between the performance of PRS at detecting breast cancer in first-degree relatives (AUCfamily) vs. the reported performance of the PRS at detecting breast cancer patients in larger European populations from previous studies (AUCpopulation). Method refers to the original method that was used to derive the weights for the PRS. b The original performance of the PRS (AUCpopulation) plotted against the improvement in PRS AUCfamily after the PRS is adjusted for population. The performance improvement is positively correlated with its original performance in the overall population suggesting the adjustment process was able to magnify burden effects while suppressing population stratification. c Evaluating the effectiveness of the best-performing PRS at predicting various cancer-related family histories. The PRS performs the best at predicting the presence of breast cancer in first-degree relatives and performance decreases as the relative degree increases and the cancer type become non-breast cancer-specific. BC: breast cancer