Tinovimba-Hove/Prostate-Cancer-Prediction-Machine-Learning

This research improves prostate cancer prediction by combining PSA levels with novel PVT1 gene biomarkers (exons 4A, 4B, and 9). Using machine learning, logistic regression and support vector machines, and validation techniques, it aims to enhance diagnostic accuracy, especially for men of African ancestry. # Enhanced Prediction of Prostate Cancer Using PVT1 Biomarkers and PSA ## Research Publication papers.ssrn.com ## Introduction Prostate cancer (PCa) is the second leading cause of cancer-related deaths among men in the United States. Current screening methods primarily utilize prostate-specific antigen (PSA), which has high sensitivity for early detection but suffers from low specificity, leading to over-diagnosis and unnecessary biopsies. This study evaluates the predictive power of Plasmacytoma variant translocation 1 (PVT1) exons 4A, 4B, and 9, which are overexpressed in prostate tissues and human serum, in conjunction with PSA to improve the specificity of prostate cancer detection. ## Objective The main objectives of this study are: - To assess the predictive capabilities of PVT1 alongside PSA in detecting prostate cancer from human serum. - To improve the specificity of PSA screening methods to reduce unnecessary biopsies. ## Methodology The analysis involved the following steps: - Data Visualization: Visualizations of the distribution of serum biomarker data across all subjects, including boxplots and correlation matrices. - Single Logistic Regression: Each biomarker was individually assessed as a predictor variable to evaluate its significance in predicting PCa status. - Multiple Logistic Regression: Various pairwise combinations of biomarkers were tested to determine the optimal model for predicting PCa status.\ - Area Under the Curve (AUC): AUC was calculated to evaluate the model's ability to discriminate between positive and negative cases. A higher AUC indicates better model performance. - 10-Fold Cross-Validation: This method was employed to assess the model's robustness and to prevent overfitting. By dividing the dataset into ten subsets, the model was trained on nine and tested on the remaining one, ensuring that every data point was used for both training and validation. - Likelihood Ratio Te …