Systematic Detection of Predictive Gene Sets by Semantics-Based Selection

This repository contains Python and R code for Gene Ontology (GO)-based feature selection and machine learning analysis of gene expression data. The pipeline supports multiple omics data formats, including:- Microarray data- RNA-seq data- DNA methylation data The datasets are evaluated using machine learning models, including:- Support Vector Machine (SVM)- Logistic Regression (L1-regularized)- Random Forest Model performance is assessed using Leave-One-Sample-Out Cross-Validation (LOSO-CV), and results are reported in terms of:- confusion matrix- Accuracy (for balanced data sets) and macro-averaged accuracy (imbalanced data sets)