The art of using t-SNE for single-cell transcriptomics

By Dmitry Kobak, Philipp Berens. 2018

1,041 citations recorded by the source corpus.

Single-cell transcriptomics yields ever growing data sets containing RNA expression levels for thousands of genes from up to millions of cells. Common data analysis pipelines include a dimensionality reduction step for visualising the data in two dimensions, most frequently performed using t-distributed stochastic neighbour embedding (t-SNE). It excels at revealing local structure in high-dimensional data, but naive applications often suffer from severe shortcomings, e.g. the global structure of the data is not represented accurately. Here we describe how to circumvent such pitfalls, and develop a protocol for creating more faithful t-SNE visualisations. It includes PCA initialisation, a high learning rate, and multi-scale similarity kernels; for very large data sets, we additionally use exaggeration and downsampling-based initialisation. We use published single-cell RNA-seq data sets to demonstrate that this protocol yields superior results compared to the naive application of t-SNE. t-SNE is widely used for dimensionality reduction and visualization of high-dimensional single-cell data. Here, the authors introduce a protocol to help avoid common shortcomings of t-SNE, for example, enabling preservation of the global structure of the data.

Subfields: 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 Molecular Biology

In Representation & Self-Supervised Learning

In Computational Biology & Bioinformatics

Citations in the corpus

Interpretable Machine Learning: Fundamental Principles and 10 Grand Challenges (2021, methodology): “Recent methods mitigate the loss of global information by using global-aware initialization (McInnes et al., 2018; Kobak and Berens, 2019) (that is, initializing the distances in the low-dimensional space using PCA) and/or selectively preserving distances between non-neighbor samples (Fu et al.,…”
Understanding How Dimension Reduction Tools Work: An Empirical Approach to Deciphering t-SNE, UMAP, TriMAP, and PaCMAP for Data Visualization (2020, background): “However, just as much recent research has discovered (Kobak and Linderman, 2019; Kobak and Berens, 2019), we found that the initialization has a much larger influence on the success rate than simply faster convergence and stability.”
Automated optimized parameters for T-distributed stochastic neighbor embedding improve visualization and analysis of large datasets (2019, methodology): “Alternatively, a ‘ late exaggeration ’ approach 21,38 can be cautiously applied to create very tight clusters, although in our experience this approach was only marginally bene ﬁ cial for global structure representation while detrimental for local structure (Suppl.”
Fast Interpolation-based t-SNE for Improved Visualization of Single-Cell RNA-Seq Data (2017, methodology): “In another very recent work, Kobak and Berens (2018) 9 proposed a protocol for exploratory analysis of scRNA-seq data using FIt-SNE (including suggested parameter choices), which leads to dramatically improved embedding quality, particularly with regard to preservation of multi-scale and global…”
UMAP does not preserve global structure any better than t-SNE when using the same initialization (2019, methodology): “Very similar embeddings can be produced by t-SNE with so called exaggeration that increases the attractive forces by a constant factor [6].”
A Unifying Perspective on Neighbor Embeddings along the Attraction-Repulsion Spectrum (2020, methodology): “singlecell biology, where researchers routinely use neighbor embedding algorithms for data visualization [2, 16].”
Dtour: a Steerable Tour De Vis through High-dimensional Data (2026, unknown): “Despite many debates about the usefulness and faithfulness of non-linear DR methods (e.g. [10, 19]), such methods are widely used and undoubtedly useful if interpreted with care [44, 17, 3].”