Polygenic risk scores for Crohn's disease risk prediction in a Chinese population: insights from multi-ethnic genome-wide association studies.

Crohn's disease (CD) is a chronic inflammatory bowel disease with a complex etiology involving genetic, immune, microbial, and environmental factors. Despite advances in understanding its pathogenesis, accurately predicting CD risk remains challenging, particularly in East Asian populations. In this study, we evaluated the performance of a polygenic risk score (PRS) model to predict CD risk in a Chinese cohort comprising whole-exome sequencing data of 76 CD patients and 552 healthy controls. We calculated PRS by applying causal genetic effect estimated from summary statistics of a public large-scale multi-ethnic genome-wide association study. Our results demonstrated that the PRS model effectively distinguished CD patients from healthy controls, achieving an area under the receiver opperating characteristic curve of 0.75 and an odds ratio of 13 for individuals with a PRS above 2.3. The model also showed consistent performance in independent control data sets of Chinese, East Asian, European, and American ancestries. These findings highlight the potential of PRS derived from multi-ethnic causal effect as a non-invasive tool for CD risk prediction in East Asian populations. However, the moderate predictive accuracy and unexplained variance emphasize the need for larger studies and the integration of additional genetic and environmental factors to refine PRS model further.