Fig. 4: Trans-ancestry PRS-CSx improved predictive performance and discrimination ability for KSD risk. | Nature Communications

Fig. 4: Trans-ancestry PRS-CSx improved predictive performance and discrimination ability for KSD risk.

From: Trans-ancestry GWAS identifies 59 loci and improves risk prediction and fine-mapping for kidney stone disease

Fig. 4

A Comparison of pseudo-R2 and AUCs for cov-only model and four PRS models. B Comparison of KSD risk by four PRSs. The central points and error bars represented ORs and the 95% confidence intervals, respectively. Source data are provided as a Source Data file. C Predictive performance of PRS in relation to known non-genetic KSD related factors. The central points and error bars represented AUC values and 95% confidence intervals, respectively. Source data are provided as a Source Data file. D Cumulative hazard curve of KSD according to stratified PRS-CSxEAS&EUR. The models constructed for Figure A and Figure B were adjusted for age, sex, and PC1-10, with an additional adjustment for non-genetic KSD-related factors for Figure D. All analyses were completed in testing dataset, including 4996 KSD cases and 199,409 controls. AUC area under receiver operating curve, cov-only covariates-only prediction model, OR odd ratio, HDL high-density lipoprotein, TG triglycerides. The glucose referred to the fasting glucose, and the calcium, urate, and vitamin D referred to the blood calcium, urate and vitamin D concentrations.

Back to article page