Dimensionality reduction by UMAP reinforces sample heterogeneity analysis in bulk transcriptomic data

• Published in: Cell reports (Cambridge) Vol. 36; no. 4; p. 109442
• Main Authors: Yang, Yang, Sun, Hongjian, Zhang, Yu, Zhang, Tiefu, Gong, Jialei, Wei, Yunbo, Duan, Yong-Gang, Shu, Minglei, Yang, Yuchen, Wu, Di, Yu, Di
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 27-07-2021; Elsevier
• Subjects: Algorithms; bulk transcriptomics; Cluster Analysis; clustering structure; Data Analysis; Databases, Genetic; dimensionality reduction; Gene Expression Profiling; heterogeneity analysis; Humans; PCA; Principal Component Analysis; Reproducibility of Results; t-SNE; UMAP; bulk transcriptomics; UMAP; dimensionality reduction; heterogeneity analysis; t-SNE; clustering structure; PCA
• ISSN: 2211-1247; 2211-1247
• DOI: 10.1016/j.celrep.2021.109442

Transcriptomic analysis plays a key role in biomedical research. Linear dimensionality reduction methods, especially principal-component analysis (PCA), are widely used in detecting sample-to-sample heterogeneity, while recently developed non-linear methods, such as t-distributed stochastic neighbor embedding (t-SNE) and uniform manifold approximation and projection (UMAP), can efficiently cluster heterogeneous samples in single-cell RNA sequencing analysis. Yet, the application of t-SNE and UMAP in bulk transcriptomic analysis and comparison with conventional methods have not been achieved. We compare four major dimensionality reduction methods (PCA, multidimensional scaling [MDS], t-SNE, and UMAP) in analyzing 71 large bulk transcriptomic datasets. UMAP is superior to PCA and MDS but shows some advantages over t-SNE in differentiating batch effects, identifying pre-defined biological groups, and revealing in-depth clusters in two-dimensional space. Importantly, UMAP generates sample clusters uncovering biological features and clinical meaning. We recommend deploying UMAP in visualizing and analyzing sizable bulk transcriptomic datasets to reinforce sample heterogeneity analysis.
[Display omitted]
•Four methods, PCA, MDS, t-SNE, and UMAP, are evaluated on 71 bulk transcriptomic datasets•UMAP is overall superior to PCA and MDS and shows some advantages over t-SNE•UMAP can efficiently and effectively reveal clusters in two-dimensional space•Clusters revealed by UMAP are associated with biological features and clinical traits
Yang et al. compare four major dimensionality reduction methods (PCA, MDS, t-SNE, and UMAP) in analyzing large bulk transcriptomic datasets. UMAP is overall superior to PCA and MDS and shows some advantages over t-SNE in differentiating batch effects, identifying pre-defined biological groups, and revealing in-depth clusters in two-dimensional space.