Geometric deep learning and multiple-instance learning for 3D cell-shape profiling
Cell Systems · 2025-03 · 16 (3) · 101229
DOI: 10.1016/j.cels.2025.101229 ↗Kurz gefasst
Kommentar in der Zeitschrift Cell Systems, 2025-03. Verfasst von Matt De Vries und 10 weiteren Autor:innen. Beteiligte Einrichtung: Department of Cancer Biology, Institute of Cancer Research, London, UK; Department of Physics, Imperial College London, London, UK; Sentinal4D, Londo…. Hauptthemen laut MEDLINE: Cell Shape, Multiple-Instance Learning Algorithms. Die Arbeit wurde bisher 24-mal zitiert (Crossref) und nennt 143 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Engineering and Physical Sciences Research Council, Cancer Research UK, National Institute for Health and Care Research.
Abstract (Originalsprache)
The three-dimensional (3D) morphology of cells emerges from complex cellular and environmental interactions, serving as an indicator of cell state and function. In this study, we used deep learning to discover morphology representations and understand cell states. This study introduced MorphoMIL, a computational pipeline combining geometric deep learning and attention-based multiple-instance learning to profile 3D cell and nuclear shapes. We used 3D point-cloud input and captured morphological signatures at single-cell and population levels, accounting for phenotypic heterogeneity. We applied these methods to over 95,000 melanoma cells treated with clinically relevant and cytoskeleton-modulating chemical and genetic perturbations. The pipeline accurately predicted drug perturbations and cell states. Our framework revealed subtle morphological changes associated with perturbations, key shapes correlating with signaling activity, and interpretable insights into cell-state heterogeneity. MorphoMIL demonstrated superior performance and generalized across diverse datasets, paving the way for scalable, high-throughput morphological profiling in drug discovery. A record of this paper's transparent peer review process is included in the supplemental information.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Cancer Biology, Institute of Cancer Research, London, UK; Department of Physics, Imperial College London, London, UK; Sentinal4D, London, UK.
- Department of Cancer Biology, Institute of Cancer Research, London, UK.
- Department of Physics, Imperial College London, London, UK.
- Gatsby Computational Neuroscience Unit, University College London, London, UK.
- Department of Cancer Biology, Institute of Cancer Research, London, UK; Sentinal4D, London, UK. Electronic address: chris.bakal@icr.ac.uk.
Förderung
- Engineering and Physical Sciences Research Council
- Cancer Research UK
- National Institute for Health and Care Research
- Institute of Cancer Research
- Commonwealth Scholarship Commission
Zitieren
APA
De Vries, M., Dent, L. G., Curry, N., Rowe-Brown, L., Bousgouni, V., Fourkioti, O., Naidoo, R., Sparks, H., Tyson, A., Dunsby, C., & Bakal, C. (2025). Geometric deep learning and multiple-instance learning for 3D cell-shape profiling. Cell Systems, 16(3), 101229. https://doi.org/10.1016/j.cels.2025.101229
Vancouver
De Vries M, Dent LG, Curry N, Rowe-Brown L, Bousgouni V, Fourkioti O, et al. Geometric deep learning and multiple-instance learning for 3D cell-shape profiling. Cell Systems. 2025;16(3):101229. doi:10.1016/j.cels.2025.101229
BibTeX
@article{DeVries2025,
author = {De Vries, Matt and Dent, Lucas G and Curry, Nathan and Rowe-Brown, Leo and Bousgouni, Vicky and Fourkioti, Olga and Naidoo, Reed and Sparks, Hugh and Tyson, Adam and Dunsby, Chris and Bakal, Chris},
title = {Geometric deep learning and multiple-instance learning for 3D cell-shape profiling},
journal = {Cell Systems},
year = {2025},
volume = {16},
number = {3},
pages = {101229},
doi = {10.1016/j.cels.2025.101229}
}
EUVIMED