Deciphering the Polygenic Basis of Racial Disparities in Prostate Cancer By an Integrative Analysis of Genomic and Transcriptomic Data

Abstract

Prostate cancer prevalence in African Americans (AA) is over 1.5 times the prevalence in European Americans (EA). Among over a hundred index risk SNPs for prostate cancer, only a few can be verified using the available AAs' data. Their relevance to the prevalence inequality and other racial disparities has not been fully determined. We investigated this issue by an integrative analysis of five public datasets. We categorized the datasets into two classes. The training class consisted of the datasets generated by three genome-wide association studies. The test class contained the prostate adenocarcinoma data of The Cancer Genome Atlas and the data of African and European super-populations in the 1000-Genome project. The polygenic risk scores (PRS) of test samples for cancer occurrence were calculated according to the effects of genetic variants estimated from the training samples. We obtained the following findings. Africans' PRSs are higher than Europeans' scores (P < 1 × 10−6). AA patients' PRSs are higher than EA patients' scores (P < 3×10−9). The patients with tumors presenting fusion or abnormal expression in ERG and other E26 transformation-specific (ETS) family genes have lower PRSs than the patients without such aberrations (P < 7×10−5). Five tumor progression-related genes have the expression levels being significantly correlated with PRS (FDR < 0.01). Additional simulation analysis shows that the high prostate cancer prevalence in African populations makes it challenging to identify individual risk variants using African men's data. These results implicate that the index risk SNP-based PRS is compatible with the observed racial disparity in prostate cancer prevalence and ETS abnormal cancers may be less heritable compared with other subtypes.

Prevention Relevance:

This study reveals the relevance of index risk SNP markers with racial disparities in prostate cancer. The findings also indicate that PRS can be used in prostate cancer subtype prediction.