EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

An interpretable and adaptive autoencoder for efficient tissue deconvolution

Jesús de la Fuente iD, Naroa Legarra-Marcos, Aintzane Diaz-Mazkiaran, Guillermo Serrano, Irene Marín-Goñi iD, Markel Benito Sendin, Ana García Osta, Krishna R Kalari, Carlos Fernandez-Granda, Idoia Ochoa, Mikel Hernaez iD

Nucleic Acids Research · 2025

DOI: 10.1093/nar/gkaf821 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Nucleic Acids Research, 2025. Verfasst von Jesús de la Fuente und 10 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biomedical Engineering and Science, Tecnun School of Engineering, University of Navarra, 20018 San Sebastian, Spain. Hauptthemen laut MEDLINE: Single-Cell Analysis, Gene Expression Profiling, Software, RNA-Seq. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Instituto de Salud Carlos III, AEI, NextGenerationEU.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Deconvolution models are a powerful tool for extracting cell-type-specific information from bulk gene expression profiles. Current methods leverage advanced machine learning models and high-resolution sequencing, like single-cell RNA-sequencing, showing promising results across diverse tissues and conditions. However, they still present important limitations: First, many depend on selecting a robust reference, which can strongly affect the deconvolution. Second, pseudobulk data used for training and real bulk RNA-seq samples often exhibit strong distribution shifts, which are currently unaccounted for. Finally, most deconvolution approaches behave as black boxes, which can compromise the reliability of the results. Here, we present Sweetwater, an adaptive and interpretable autoencoder that efficiently deconvolves bulk samples leveraging multiple classes of reference data. Moreover, we propose an improved way of generating training data from a mixture of FACS-sorted FASTQ files, reducing platform-specific biases and outperforming current single-cell-based references. Furthermore, we introduce a gold standard dataset to facilitate fair and accurate evaluation of deconvolution approaches. Finally, we demonstrate that Sweetwater adapts effectively to deconvolved samples during training, uncovering biologically meaningful patterns and enhancing result’s reliability. Sweetwater is available at https://doi.org/10.6084/m9.figshare.29609180, and we anticipate it will expedite the accurate examination of high-throughput clinical data across diverse applications.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Single-Cell AnalysisGene Expression ProfilingSoftwareRNA-SeqHumansMachine LearningAlgorithmsAnimalsSequence Analysis, RNAMiceTranscriptomeOrgan SpecificityAutoencoder

Autor:innen und Institutionen

  • Department of Biomedical Engineering and Science, Tecnun School of Engineering, University of Navarra, 20018 San Sebastian, Spain.
  • CIMA, University of Nava rra, IdiSNA, 31008 Pamplona, Spain.
  • Biological & Environmental Science and Engineering, KAUST, 23955 Thuwal, Saudi Arabia.
  • Department of Molecular Pharmacology, Mayo Clinic, 55905 Minnesota, United States.
  • Department of Quantitative Health Sciences, Mayo Clinic, 55905 Minnesota, United States.
  • Center for Data Science, New York University, 10011 New York, United States.

ORCID-Profile: Jesús de la Fuente, Irene Marín-Goñi, Mikel Hernaez

Förderung

  • Instituto de Salud Carlos III
  • AEI
  • NextGenerationEU
  • Fulbright Predoctoral Research Program
  • Fundacion Ramon Areces
  • Gipuzkoa Fellows
  • Basque Government
  • Ministry of Science and Innovation

Zitieren

APA

de la Fuente, J., Legarra-Marcos, N., Diaz-Mazkiaran, A., Serrano, G., Marín-Goñi, I., Sendin, M. B., Osta, A. G., Kalari, K. R., Fernandez-Granda, C., Ochoa, I., & Hernaez, M. (2025). An interpretable and adaptive autoencoder for efficient tissue deconvolution. Nucleic Acids Research. https://doi.org/10.1093/nar/gkaf821

Vancouver

de la Fuente J, Legarra-Marcos N, Diaz-Mazkiaran A, Serrano G, Marín-Goñi I, Sendin MB, et al. An interpretable and adaptive autoencoder for efficient tissue deconvolution. Nucleic Acids Research. 2025. doi:10.1093/nar/gkaf821

BibTeX
@article{delaFuente2025,
  author = {de la Fuente, Jesús and Legarra-Marcos, Naroa and Diaz-Mazkiaran, Aintzane and Serrano, Guillermo and Marín-Goñi, Irene and Sendin, Markel Benito and Osta, Ana García and Kalari, Krishna R and Fernandez-Granda, Carlos and Ochoa, Idoia and Hernaez, Mikel},
  title = {An interpretable and adaptive autoencoder for efficient tissue deconvolution},
  journal = {Nucleic Acids Research},
  year = {2025},
  doi = {10.1093/nar/gkaf821}
}
Kostenlos starten →