School of Medicine, Alma Mater Studiorum, Università Di Bologna, Via Zamboni, 33, 40126, Bologna, Italy.
Department of Agroecology, College of Agriculture and Natural Resources of Darab, Shiraz University, Shiraz, Iran.
World J Surg Oncol. 2024 Jul 5;22(1):177. doi: 10.1186/s12957-024-03459-6.
This study investigates the genetic factors contributing to the disparity in prostate cancer incidence and progression among African American men (AAM) compared to European American men (EAM). The research focuses on employing Weighted Gene Co-expression Network Analysis (WGCNA) on public microarray data obtained from prostate cancer patients. The study employed WGCNA to identify clusters of genes with correlated expression patterns, which were then analyzed for their connection to population backgrounds. Additionally, pathway enrichment analysis was conducted to understand the significance of the identified gene modules in prostate cancer pathways. The Least Absolute Shrinkage and Selection Operator (LASSO) and Correlation-based Feature Selection (CFS) methods were utilized for selection of biomarker genes. The results revealed 353 differentially expressed genes (DEGs) between AAM and EAM. Six significant gene expression modules were identified through WGCNA, showing varying degrees of correlation with prostate cancer. LASSO and CFS methods pinpointed critical genes, as well as six common genes between both approaches, which are indicative of their vital role in the disease. The XGBoost classifier validated these findings, achieving satisfactory prediction accuracy. Genes such as APRT, CCL2, BEX2, MGC26963, and PLAU were identified as key genes significantly associated with cancer progression. In conclusion, the research underlines the importance of incorporating AAM and EAM population diversity in genomic studies, particularly in cancer research. In addition, the study highlights the effectiveness of integrating machine learning techniques with gene expression analysis as a robust methodology for identifying critical genes in cancer research.
本研究旨在探讨导致非裔美国男性（AAM）与欧洲裔美国男性（EAM）之间前列腺癌发病率和进展差异的遗传因素。研究采用加权基因共表达网络分析（WGCNA）方法，对前列腺癌患者的公共微阵列数据进行分析。该研究运用 WGCNA 方法识别具有相关表达模式的基因簇，然后分析它们与人群背景的关系。此外，还进行了通路富集分析，以了解鉴定基因模块在前列腺癌通路中的意义。采用最小绝对收缩和选择算子（LASSO）和基于相关性的特征选择（CFS）方法选择生物标志物基因。研究结果显示，AAM 和 EAM 之间存在 353 个差异表达基因（DEGs）。通过 WGCNA 鉴定出 6 个显著的基因表达模块，它们与前列腺癌的相关性程度不同。LASSO 和 CFS 方法确定了关键基因，以及两种方法之间的 6 个共同基因，表明它们在疾病中具有重要作用。XGBoost 分类器验证了这些发现，达到了令人满意的预测准确性。鉴定出 APRT、CCL2、BEX2、MGC26963 和 PLAU 等关键基因，它们与癌症进展显著相关。综上所述，该研究强调了在基因组研究中纳入 AAM 和 EAM 人群多样性的重要性，特别是在癌症研究中。此外，该研究还强调了将机器学习技术与基因表达分析相结合作为一种识别癌症研究中关键基因的有效方法的重要性。
