Prostate Cancer Biomarker Identification using Machine learning approach

Prostate cancer (PCa) is the most common cancer in men (20%) and accounts for one-fifth (6.8%) of cancer-related deaths in men worldwide. The most prevalent risk factors for PCa include smoking, obesity, race/ethnicity, food, age, chemical and radiation exposure, sexually transmitted illnesses, and so on. The fundamental shift at the molecular level, however, is the confirmation of PCa. Therefore, in this work we uses the DESeq2 Analysis in R and the XGBoost Machine Learning technique to describe differentially expressed genes (DEG). After analysing all our datasets, we discovered 6868 most significant upregulated genes and 6338 most significant downregulated genes, totaling 13206 DEGs. most significant upregulated genes and 1219 downregulated genes, and a sum total of 1852 DEGs were found from all four datasets after assessment. The significance criteria was set at a p adjusted value of less than 0.05 and a log2fold value of more than 2. The genes ANGPT1, APOBEC3C, ZNF185, EPHA10, and HOXC4 have the potential to be used as diagnostic tests for prostate cancer early detection. Our main goal in this project was to uncover several biomarkers that may be explored for efficient prostate cancer research and how these biomarkers can be used to detect prostate cancer. Another key reason we need to conduct research on this topic is because these tests are highly expensive to conduct, and so cannot benefit a larger segment of our country's population who cannot afford them.