Refining cancer prediction with DNA sequencing and combined machine learning approaches

A high-accuracy DNA-based cancer risk predictor was developed by blending Logistic Regression with Gaussian Naive Bayes, and its hyperparameters were optimized via grid search. Five cancer types (BRCA1, KIRC, COAD, LUAD, PRAD) were classified in a cohort of 390 patients. Accuracies of 100% were attained for BRCA1, KIRC, and COAD, while 98% was achieved for LUAD and PRAD-representing improvements of 1-2% over recent deep-learning and multi-omic benchmarks. A micro- and macro-average ROC AUC of 0.99 was obtained. The blended ensemble was shown to outperform each individual algorithm and surpass existing state-of-the-art methods, providing a lightweight, interpretable, and highly effective tool for early cancer prediction.