Lifting the curse from high-dimensional data: automated projection pursuit clustering for a variety of biological data modalities
Unsupervised clustering is a powerful machine-learning technique widely used to analyze high-dimensional biological data. It plays a crucial role in uncovering patterns, structures, and inherent relationships within complex datasets without relying on predefined labels. In the context of biology, high-dimensional data may include transcriptomics, proteomics, and a variety of single-cell omics data. Most existing clustering algorithms operate directly in the high-dimensional space, and their performance may be negatively affected by the phenomenon known as the curse of dimensionality. Here, we show an alternative clustering approach that alleviates the curse by sequentially projecting high-dimensional data into a low-dimensional representation. We validated the effectiveness of our approach, named automated projection pursuit (APP), across various biological data modalities, including flow and mass cytometry data, scRNA-seq, multiplex imaging data, and T-cell receptor repertoire data. APP efficiently recapitulated experimentally validated cell-type definitions and revealed new biologically meaningful patterns.
Authors
- Eliver Ghosn (ORCID: https://orcid.org/0000-0001-7258-906X)
- Connor Meehan (ORCID: https://orcid.org/0000-0002-7596-2437)
- Guenther Walther (ORCID: https://orcid.org/0000-0002-0374-123X)
- C. Schiller (ORCID: https://orcid.org/0009-0000-3771-2202)
- Evgeniy Tabatsky
- Florian Georgescauld (ORCID: https://orcid.org/0000-0003-0530-1884)
- Devon J. Eddins (ORCID: https://orcid.org/0000-0002-7905-8306)
- Tyler Levy (ORCID: https://orcid.org/0000-0003-4634-4063)
- Claire Simpson
- Denis Schapiro (ORCID: https://orcid.org/0000-0002-9391-5722)
- Sasha Tkachev (ORCID: https://orcid.org/0000-0003-3984-8351)
- Kresimir Bestak (ORCID: https://orcid.org/0009-0009-8245-9846)
- Derek Papalegis
- Zainab Rahil
- Andrei V. Chernyshev (ORCID: https://orcid.org/0000-0002-7224-1992)
- Martin Culka
- Darya Orlova
- Ivan V. Gregoretti (ORCID: https://orcid.org/0000-0001-6965-2494)
Institutions
- Emory University (US)
- Heidelberg University (DE)
- University Hospital Heidelberg (DE)
- Cell Signaling Technology (United States) (US)
- Komsomolsk-on-Amur State Technical University (RU)
- Institute of Chemical Kinetics and Combustion (RU)
- Columbia University (US)
- Stanford University (US)
Publication Details
- Journal
- GigaScience
- Published
- 2025-01-01
- DOI
- https://doi.org/10.1093/gigascience/giaf052
- Citations
- 4
- Primary Topic
- Single-cell and spatial transcriptomics
- Type
- article
- Field-Weighted Citation Impact
- 2.42
Funders
- Emory University
- Deutsche Forschungsgemeinschaft
- Bundesministerium für Bildung und Forschung
- National Institutes of Health
- National Institute of Allergy and Infectious Diseases