Dimensionality reduction by UMAP reinforces sample heterogeneity analysis in bulk transcriptomic data
Summary: 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...
| Main Authors: | , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2021-07-01
|
| Series: | Cell Reports |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2211124721008597 |
| _version_ | 1819131666740805632 |
|---|---|
| author | Yang Yang Hongjian Sun Yu Zhang Tiefu Zhang Jialei Gong Yunbo Wei Yong-Gang Duan Minglei Shu Yuchen Yang Di Wu Di Yu |
| author_facet | Yang Yang Hongjian Sun Yu Zhang Tiefu Zhang Jialei Gong Yunbo Wei Yong-Gang Duan Minglei Shu Yuchen Yang Di Wu Di Yu |
| author_sort | Yang Yang |
| collection | DOAJ |
| description | Summary: 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. |
| first_indexed | 2024-12-22T09:19:08Z |
| format | Article |
| id | doaj.art-260c025e813f47bfa2ae65a54b9ed7af |
| institution | Directory Open Access Journal |
| issn | 2211-1247 |
| language | English |
| last_indexed | 2024-12-22T09:19:08Z |
| publishDate | 2021-07-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Cell Reports |
| spelling | doaj.art-260c025e813f47bfa2ae65a54b9ed7af2022-12-21T18:31:14ZengElsevierCell Reports2211-12472021-07-01364109442Dimensionality reduction by UMAP reinforces sample heterogeneity analysis in bulk transcriptomic dataYang Yang0Hongjian Sun1Yu Zhang2Tiefu Zhang3Jialei Gong4Yunbo Wei5Yong-Gang Duan6Minglei Shu7Yuchen Yang8Di Wu9Di Yu10The University of Queensland Diamantina Institute, Faculty of Medicine, The University of Queensland, Translational Research Institute, Brisbane, QLD, Australia; Shandong Artificial Intelligence Institute, Qilu University of Technology (Shandong Academy of Sciences), Jinan, ChinaShandong Artificial Intelligence Institute, Qilu University of Technology (Shandong Academy of Sciences), Jinan, China; School of Microelectronics, Shandong University, Jinan, ChinaLaboratory of Immunology for Environment and Health, School of Pharmaceutical Sciences, Qilu University of Technology (Shandong Academy of Sciences), Jinan, ChinaUniversity of Electronic Science and Technology of China, Chengdu, ChinaShenzhen Key Laboratory of Fertility Regulation, Center of Assisted Reproduction and Embryology, University of Hong Kong, Shenzhen Hospital, Shenzhen, ChinaLaboratory of Immunology for Environment and Health, School of Pharmaceutical Sciences, Qilu University of Technology (Shandong Academy of Sciences), Jinan, ChinaShenzhen Key Laboratory of Fertility Regulation, Center of Assisted Reproduction and Embryology, University of Hong Kong, Shenzhen Hospital, Shenzhen, ChinaShandong Artificial Intelligence Institute, Qilu University of Technology (Shandong Academy of Sciences), Jinan, ChinaDepartment of Pathology and Laboratory Medicine, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA; McAllister Heart Institute, University of North Carolina at Chapel Hill, Chapel Hill, NC, USADepartment of Biostatistics, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA; Division of Oral and Craniofacial Health Science, Adams School of Dentistry, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA; Lineberger Comprehensive Cancer Center, University of North Carolina, Chapel Hill, NC, USA; Corresponding authorThe University of Queensland Diamantina Institute, Faculty of Medicine, The University of Queensland, Translational Research Institute, Brisbane, QLD, Australia; Shandong Artificial Intelligence Institute, Qilu University of Technology (Shandong Academy of Sciences), Jinan, China; Laboratory of Immunology for Environment and Health, School of Pharmaceutical Sciences, Qilu University of Technology (Shandong Academy of Sciences), Jinan, China; Corresponding authorSummary: 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.http://www.sciencedirect.com/science/article/pii/S2211124721008597bulk transcriptomicsdimensionality reductionUMAPt-SNEPCAclustering structure |
| spellingShingle | Yang Yang Hongjian Sun Yu Zhang Tiefu Zhang Jialei Gong Yunbo Wei Yong-Gang Duan Minglei Shu Yuchen Yang Di Wu Di Yu Dimensionality reduction by UMAP reinforces sample heterogeneity analysis in bulk transcriptomic data Cell Reports bulk transcriptomics dimensionality reduction UMAP t-SNE PCA clustering structure |
| title | Dimensionality reduction by UMAP reinforces sample heterogeneity analysis in bulk transcriptomic data |
| title_full | Dimensionality reduction by UMAP reinforces sample heterogeneity analysis in bulk transcriptomic data |
| title_fullStr | Dimensionality reduction by UMAP reinforces sample heterogeneity analysis in bulk transcriptomic data |
| title_full_unstemmed | Dimensionality reduction by UMAP reinforces sample heterogeneity analysis in bulk transcriptomic data |
| title_short | Dimensionality reduction by UMAP reinforces sample heterogeneity analysis in bulk transcriptomic data |
| title_sort | dimensionality reduction by umap reinforces sample heterogeneity analysis in bulk transcriptomic data |
| topic | bulk transcriptomics dimensionality reduction UMAP t-SNE PCA clustering structure |
| url | http://www.sciencedirect.com/science/article/pii/S2211124721008597 |
| work_keys_str_mv | AT yangyang dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT hongjiansun dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT yuzhang dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT tiefuzhang dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT jialeigong dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT yunbowei dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT yonggangduan dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT mingleishu dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT yuchenyang dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT diwu dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata AT diyu dimensionalityreductionbyumapreinforcessampleheterogeneityanalysisinbulktranscriptomicdata |