A deep generative decoder predicts microRNA expression from bulk and single-cell mRNA profiles
- 1Department of Molecular Medicine, Aarhus University Hospital, Denmark
Abstract
MicroRNAs (miRNAs) are key post-transcriptional regulators, yet standard bulk and single-cell RNA-seq do not capture them, leaving this regulatory layer invisible in most transcriptomic data. We present miDGD, a deep generative decoder that jointly models paired mRNA and miRNA profiles through a shared latent representation, enabling miRNA expression to be predicted from mRNA alone. Trained on tumors (TCGA), healthy tissues (GTEx), and human cell lines, miDGD recovers hundreds of miRNAs in held-out tumors (mean Spearman {rho} = 0.56), captures both tissue-specific and ubiquitous miRNAs, and preserves known miRNA--target repression and host-gene co-expression. Without label supervision, its latent space separates 32 cancer types (80% accuracy). Predictions remain stable at single-cell-like sparsity and transfer across datasets--from tumors to healthy tissues and from bulk to single cells--where miDGD outperforms existing supervised and activity-inference methods. miDGD thus unlocks miRNA regulation in the vast body of existing mRNA-only data, including single-cell datasets.
Related
- MicroRNA dysregulation in pemphigus: a systematic review of expression profiles and molecular targets
- Gene and microRNA Expression Profiling in Experimental Colitis and Human Ulcerative Colitis: Effects of STAT6 Knockout and Corylus avellana Treatment
- Social isolation is associated with microRNA expression changes in the visual system of poison frog tadpoles
- The increased production of monocytes with immunosuppressive transcriptional profiles predicts the development of SIV encephalitis
- Interaction Profiles as a Universal Language for Generative Molecular Design with ShEPhERD-2
BibTeX
@misc{zamani2026deep,
title = {{A deep generative decoder predicts microRNA expression from bulk and single-cell mRNA profiles}},
author = {F. Zamani and A. M. Rasmussen and V. Schuster and M. H. Diekema and A. Krogh and J. S. Pedersen},
howpublished = {bioRxiv},
year = {2026},
month = may,
doi = {10.64898/2026.05.29.727918},
url = {https://www.biorxiv.org/content/10.64898/2026.05.29.727918v2},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - A deep generative decoder predicts microRNA expression from bulk and single-cell mRNA profiles
AU - F. Zamani
AU - A. M. Rasmussen
AU - V. Schuster
AU - M. H. Diekema
AU - A. Krogh
AU - J. S. Pedersen
PY - 2026
DA - 2026/05/29
PB - bioRxiv
DO - 10.64898/2026.05.29.727918
UR - https://www.biorxiv.org/content/10.64898/2026.05.29.727918v2
AB - MicroRNAs (miRNAs) are key post-transcriptional regulators, yet standard bulk and single-cell RNA-seq do not capture them, leaving this regulatory layer invisible in most transcriptomic data. We present miDGD, a deep generative decoder that jointly models paired mRNA and miRNA profiles through a shared latent representation, enabling miRNA expression to be predicted from mRNA alone. Trained on tumors (TCGA), healthy tissues (GTEx), and human cell lines, miDGD recovers hundreds of miRNAs in held-out tumors (mean Spearman {rho} = 0.56), captures both tissue-specific and ubiquitous miRNAs, and preserves known miRNA--target repression and host-gene co-expression. Without label supervision, its latent space separates 32 cancer types (80% accuracy). Predictions remain stable at single-cell-like sparsity and transfer across datasets--from tumors to healthy tissues and from bulk to single cells--where miDGD outperforms existing supervised and activity-inference methods. miDGD thus unlocks miRNA regulation in the vast body of existing mRNA-only data, including single-cell datasets.
ER -