Cross-modal mapping of cancer stem-like cell plasticity using deep learning
NAR Cancer · 27.07.2026 · 8 (3)
DOI: 10.1093/narcan/zcag015 ↗Kurz gefasst
Fachartikel in der Zeitschrift NAR Cancer, online erschienen am 27.07.2026. Verfasst von Debojyoti Chowdhury und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Chemical and Biological Sciences, S.N. Bose National Centre for Basic Sciences, Kolkata 700106, India. Hauptthemen laut MEDLINE: Neoplastic Stem Cells, Deep Learning, Neoplasms, Cell Plasticity. Die Arbeit nennt 96 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Mphasis F1 Foundation, DBT-JRF Fellowship, ANRF-ARG.
Abstract (Originalsprache)
Cancer stem-like cells (CSCs) play a pivotal role in driving tumor heterogeneity, therapeutic resistance, and disease progression. Despite the power of single-cell RNA sequencing (scRNA-seq) to resolve intratumoral hierarchies, there remains a need for robust, scalable tools to consistently profile CSCs across both single-cell and bulk transcriptomic data. To address this, we developed ACSCeND—a unified, machine learning–based framework that enables high-resolution CSC state classification and tissue-level deconvolution. ACSCeND comprises (i) a supervised classifier trained on curated scRNA-seq datasets to assign cells into pluripotent-like, multipotent-like, or unipotent-like states, and (ii) an attention-guided autoencoder that deconvolves CSC subtype proportions from bulk RNA sequencing data. Compared to existing tissue deconvolution tools, ACSCeND achieves superior performance, with higher accuracy across synthetic and real-world samples. Applied to over 25 000 tumor profiles from The Cancer Genome Atlas (TCGA), PREdiction of Clinical Outcomes from Genomics (PRECOG), tumor-relapse, and checkpoint inhibitor studies, ACSCeND reveals that CSC abundance strongly correlates with poor disease-free survival and reduced immunotherapy efficacy. Moreover, it uncovers distinct CSC-state-specific molecular programs, offering insights into CSC-driven heterogeneity and tumor evolution. The model also recapitulates known developmental hierarchies in noncancerous tissues, supporting its broader biological relevance. By integrating single-cell precision with bulk-level applicability, ACSCeND offers a robust, interpretable approach to profiling CSC dynamics and establishes CSC state as a clinically meaningful, pan-cancer biomarker for guiding stemness-informed therapies. ACSCeND is available as a python package (through pip) at https://pypi.org/project/ACSCeND/.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- Department of Chemical and Biological Sciences, S.N. Bose National Centre for Basic Sciences, Kolkata 700106, India.
- Department of Computer Science, Ashoka University, Sonipat, Haryana 131029, India.
- Technical Research Centre, S.N. Bose National Centre for Basic Sciences, Kolkata 700106, India.
ORCID-Profile: Bhavesh Neekhra, Shubhasis Haldar
Förderung
- Mphasis F1 Foundation
- DBT-JRF Fellowship
- ANRF-ARG
Zitieren
APA
Chowdhury, D., Priyadarshi, S., Biswas, S., Neekhra, B., Gupta, D., & Haldar, S. (2026). Cross-modal mapping of cancer stem-like cell plasticity using deep learning. NAR Cancer, 8(3). https://doi.org/10.1093/narcan/zcag015
Vancouver
Chowdhury D, Priyadarshi S, Biswas S, Neekhra B, Gupta D, Haldar S. Cross-modal mapping of cancer stem-like cell plasticity using deep learning. NAR Cancer. 2026;8(3). doi:10.1093/narcan/zcag015
BibTeX
@article{Chowdhury2026,
author = {Chowdhury, Debojyoti and Priyadarshi, Shreyansh and Biswas, Sayan and Neekhra, Bhavesh and Gupta, Debayan and Haldar, Shubhasis},
title = {Cross-modal mapping of cancer stem-like cell plasticity using deep learning},
journal = {NAR Cancer},
year = {2026},
volume = {8},
number = {3},
doi = {10.1093/narcan/zcag015}
}
EUVIMED