The art of using t-SNE for single-cell transcriptomics
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.
Introduction Recent years have seen a rapid growth of interest in single-cell RNA sequencing (scRNA-seq), or single-cell transcriptomics1,2. Through improved experimental techniques it has become possible to obtain gene expression data from thousands or even millions of cells3,4,5,6,7,8. Computational analysis of such data sets often entails unsupervised, exploratory steps including dimensionality reduction for visualisation. To this end, many studies today are using t-distributed stochastic neighbour embedding, or t-SNE9. This technique maps a set of high-dimensional points to two…
saved by
related reading
- How to Use t-SNE Effectivelydistill.pub
- A pan-cancer somatic mutation embedding using autoencodersbmcbioinformatics.biomedcentral.com
- Visualizing Data using t-SNEjmlr.org
- Quantifying the effect of experimental perturbations at single-cell resolutionnature.com
- Dimensionality reduction in neural data analysisxcorr.net
- 10. Feature selection — Single-cell best practicessc-best-practices.org
- Chapter 3 Feature selection | Basics of Single-Cell Analysis with Bioconductorbioconductor.org
- scGen predicts single-cell perturbation responsesnature.com
- Evaluating the role of pretraining dataset size and diversity on single-cell foundation model performancenature.com
- 2.2. Manifold learning — scikit-learn 1.9.0 documentationscikit-learn.org
- A survey of best practices for RNA-seq data analysisgenomebiology.biomedcentral.com
- Signature-scoring methods developed for bulk samples are not adequate for cancer single-cell RNA sequencing data | eLifeelifesciences.org