Synthetic Data Generation for Omics

Provides methods for generating synthetic omics data for: - Data augmentation in imbalanced datasets - Privacy-preserving data sharing - Simulation studies

Details

Methods include: - **SMOTE**: Synthetic Minority Over-sampling for imbalanced classes - **Gaussian Noise**: Add random perturbations to existing samples - **TabDDPM**: Diffusion-based generative model (requires torch)