Towards Useful and Private Synthetic Omics: Community Benchmarking of Generative Models for Transcriptomics Data

2026/03/04 by Hakime Öztürk, Tejumade Afonja, Joonas Jälkö +19 · 1 voice
Biochemistry, Genetics and Molecular Biology · Computer Science · Medicine · #Single-cell and spatial transcriptomics #Privacy-Preserving Technologies in Data #Ferroptosis and cancer prognosis

paper · doi:10.64898/2026.03.02.707794

openalex publication_date 2026/03/04 · openalex created_date 2026/03/05 · openalex updated_date 2026/07/22

Abstract

Abstract Background The synthesis of anonymized data derived from real-world cohorts offers a promising strategy for regulatory-compliant and privacy-preserving biological data sharing, potentially facilitating model development that can improve predictive performance. However, the extent to which generative models can preserve biological signals while remaining resilient to adversarial privacy attacks in high-dimensional omics contexts remains underexplored. To address this gap, the CAMDA 2025 Health Privacy Challenge launched a community-driven effort to systematically benchmark synthetic and privacy-preserving data generation for bulk RNA-seq cohorts. Results Building on this initiative, we systematically benchmarked 11 generative methods across two cancer cohorts (∼1,000 and ∼5,000 patients) over 978 landmark genes. Methods were evaluated across complementary axes of distributional fidelity, downstream utility, biological plausibility and empirical privacy risk, with emphasis on trade-offs between vulnerability to membership inference attacks (MIA) and other evaluation dimensions. Expressive deep generative models achieved strong predictive utility and differential expression recovery, but were often more vulnerable to membership inference risk. Differentially private methods improved resistance to attacks at the cost of reduced utility, while simpler statistical approaches offered competitive utility with moderate privacy risk and fast training. Conclusions Synthetic bulk RNA-seq quality is inherently multi-dimensional and shaped by trade-offs between utility, biological preservation and privacy. Our results indicate that differences in model architecture drive distinct trade-offs across these axes, suggesting that model choice should align with dataset characteristics, intended downstream use and privacy requirements. Privacy risk should also be assessed using multiple complementary attack methods and, where possible, formal differential privacy protection.

Citations

• Deep Unsupervised Learning using Nonequilibrium Thermodynamics
• The Cancer Genome Atlas Pan-Cancer analysis project
• Denoising Diffusion Probabilistic Models
• Differentially Private Diffusion Models
• TAPAS: a Toolbox for Adversarial Privacy Auditing of Synthetic Data
• Membership Inference Attacks against Synthetic Data through Overfitting Detection
• Synthetic Data: Revisiting the Privacy-Utility Trade-off
• Generative AI in Depth: A Survey of Recent Advances, Model Variants, and Real-World Applications
• A Consensus Privacy Metrics Framework for Synthetic Data
• The Inadequacy of Similarity-based Privacy Metrics: Privacy Attacks against "Truly Anonymous" Synthetic Datasets
• Generative models improve fairness of medical classifiers under distribution shifts
• SoK: Let the Privacy Games Begin! A Unified Treatment of Data Inference Privacy in Machine Learning
• Generating Synthetic Mixed-type Longitudinal Electronic Health Records for Artificial Intelligent Applications
• Membership Inference Attacks From First Principles
• Opacus: User-Friendly Differential Privacy Library in PyTorch
• On Utility and Privacy in Synthetic Genomic Data
• Differentially Private Learning Needs Better Features (or Much More Data)
• Synthetic Data -- Anonymisation Groundhog Day
• Denoising Diffusion Implicit Models
• PyTorch: An Imperative Style, High-Performance Deep Learning Library
• Modeling Tabular data using Conditional GAN
• Reconstruction and Membership Inference Attacks against Generative\n Models
• Disparate Vulnerability to Membership Inference Attacks
• Graphical-model based estimation and inference for differential privacy
• Improved Training of Wasserstein GANs
• Membership Inference Attacks against Machine Learning Models
• Federated Learning: Strategies for Improving Communication Efficiency

Discussions

• Thanks to CAMDA for providing the platform, and to the organizing team and all participants who helped make this effort possible. 📄 Our preprint is now available on bioRxiv - explore it here: doi.or [bsky, 0 points, 1 comments]