PuSH - Publication Server of Helmholtz Zentrum München

Inecik, K. ; Kara, A. ; Rose, A.* ; Haniffa, M.* ; Theis, F.J.

TarDis: Achieving robust and structured disentanglement of multiple covariates.

Cell Syst. 17:101573 (2026)
Publ. Version/Full Text Research data DOI PMC
Open Access Hybrid
Creative Commons Lizenzvertrag
Addressing challenges in domain invariance within single-cell genomics necessitates innovative strategies for managing the heterogeneity of multi-source datasets while maintaining the integrity of biological signals. We introduce targeted disentanglement (TarDis), an end-to-end deep generative model designed to disentangle intricate covariate structures across diverse biological datasets, distinguishing technical artifacts from true biological variations. By employing tailored covariate-specific loss components and a self-supervised approach, TarDis effectively generates multiple latent-space representations that capture each continuous and categorical target covariate separately, along with unexplained variation. Our extensive evaluations demonstrate that TarDis outperforms existing methods in data integration, covariate disentanglement, and robust out-of-distribution predictions. The model's capacity to produce interpretable and structured latent spaces, including its introduction of ordered latent representations for continuous covariates, markedly enhances its utility in hypothesis-driven research. Consequently, TarDis offers a promising analytical platform for advancing scientific discovery, providing insights into cellular dynamics, and enabling targeted therapeutic interventions.
Altmetric
Additional Metrics?
[➜Log in]
Edit extra informations Login
Publication type Article: Journal article
Document type Scientific Article
Keywords Continuous Covariate Modeling ; Covariate Disentanglement ; Generative Models ; Invariant Representation Learning ; Latent-space Optimization ; Multi-condition Data Integration ; Out-of-distribution Generalization ; Self-supervised Learning ; Single-cell Genomics; Cell Atlas
ISSN (print) / ISBN 2405-4712
e-ISSN 2405-4720
Journal Cell Systems
Quellenangaben Volume: 17, Issue: 5, Pages: , Article Number: 101573 Supplement: ,
Publisher Elsevier
Publishing Place Maryland Heights, MO
Institute(s) Institute of Computational Biology (ICB)
Grants Helmholtz Zentrum Munchen-Deutsches Forschungszentrum fur Gesundheit und Umwelt (GmbH)