RareDisease-Transcriptome-VAE: Generative Augmentation Framework for Rare Disease Transcriptomics

This Zenodo record provides the full data and model package associated with the study"Robust Generative Augmentation of Rare Disease Transcriptomes: Sampling Temperature Scalingto Mitigate Mode Collapse in VAEs". The archive includes:- Processed transcriptome matrices derived from publicly available GEO datasets- Synthetic transcriptome data generated using variational autoencoders (VAEs)- Trained model checkpoints- Robustness evaluation outputs from multiple independent runs All raw transcriptome data are publicly available from NCBI GEO and are not redistributed here.This archive is intended to support full reproducibility of the reported experiments.This Zenodo archive mirrors the exact directory structure of the GitHub repository for full reproducibility.