Evaluation of UMAP as an alternative to t-SNE for single-cell data

By Etienne Becht, Charles-Antoine Dutertre, I. Kwok, Lai Guan Ng, Florent Ginhoux, Evan W. Newell. 2018

52 citations recorded by the source corpus.

Uniform Manifold Approximation and Projection (UMAP) is a recently-published non-linear dimensionality reduction technique. Another such algorithm, t-SNE, has been the default method for such task in the past years. Herein we comment on the usefulness of UMAP high-dimensional cytometry and single-cell RNA sequencing, notably highlighting faster runtime and consistency, meaningful organization of cell clusters and preservation of continuums in UMAP compared to t-SNE.

Subfields: Biomedical Engineering, Biophysics, Molecular Biology, Representation & Self-Supervised Learning, Computational Biology & Bioinformatics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biophysics

1. Philipp Berens
2. Dmitry Kobak

In Molecular Biology

1. Dmitry Kobak
2. Philipp Berens
3. George C. Linderman
4. R. Bradford
5. B. Bruce

In Representation & Self-Supervised Learning

1. Dmitry Kobak
2. Philipp Berens
3. Felipe Meneguzzi
4. Odinaldo Rodrigues
5. Michael Luck
6. Nathan Gavenski
7. George C. Linderman
8. J. Böhm

In Computational Biology & Bioinformatics

1. Dmitry Kobak
2. Philipp Berens
3. George C. Linderman

Citations in the corpus

• The art of using t-SNE for single-cell transcriptomics (2018, methodology): “1.11 Comparison with other methods A very promising method called UMAP (McInnes and Healy, 2018) has recently attracted considerable attention from the transcriptomics community (Becht et al., 2018).”
• A Survey of Imitation Learning Methods, Environments and Metrics (2024, background): “…hyperparameters, as pointed out by van der Maaten and Hinton [109]; (ii) interpreting them might be challenging [113], which makes them less suitable as an explainable method; and (iii) and some of these techniques lack robustness and might not provide consistent results for similar datasets [10].”
• Imitation Learning: A Survey of Learning Methods, Environments and Metrics (2024, unknown): “…to hyperparameters, as pointed out by van der Maaten and Hinton [97]; (ii) interpreting them might be challenging [98], which makes them less suitable as an explainable method; and (iii) and some of these techniques lack robustness and might not provide consistent results for similar datasets [99].”