Rxivist logo

UMAP does not preserve global structure any better than t-SNE when using the same initialization

By Dmitry Kobak, George C. Linderman

Posted 19 Dec 2019
bioRxiv DOI: 10.1101/2019.12.19.877522

One of the most ubiquitous analysis tools employed in single-cell transcriptomics and cytometry is t-distributed stochastic neighbor embedding (t-SNE), used to visualize individual cells as points on a 2D scatter plot such that similar cells are positioned close together. Recently, a related algorithm, called uniform manifold approximation and projection (UMAP) has attracted substantial attention in the single-cell community. In Nature Biotechnology, Becht et al. argued that UMAP is preferable to t-SNE because it better preserves the global structure of the data and is more consistent across runs. Here we show that this alleged superiority of UMAP can be entirely attributed to different choices of initialization in the implementations used by Becht et al.: t-SNE implementations by default used random initialization, while the UMAP implementation used a technique called Laplacian eigenmaps to initialize the embedding. We show that UMAP with random initialization preserves global structure as poorly as t-SNE with random initialization, while t-SNE with informative initialization performs as well as UMAP with informative initialization. Hence, contrary to the claims of Becht et al., their experiments do not demonstrate any advantage of the UMAP algorithm per se, but rather warn against using random initialization.

Download data

  • Downloaded 2,891 times
  • Download rankings, all-time:
    • Site-wide: 2,105 out of 94,912
    • In bioinformatics: 371 out of 8,837
  • Year to date:
    • Site-wide: 803 out of 94,912
  • Since beginning of last month:
    • Site-wide: 1,491 out of 94,912

Altmetric data

Downloads over time

Distribution of downloads per paper, site-wide


Sign up for the Rxivist weekly newsletter! (Click here for more details.)