Spotless, a reproducible pipeline for benchmarking cell type deconvolution in spatial transcriptomics
- Author
- Chananchida Sang-aram (UGent) , Robin Browaeys (UGent) , Ruth Seurinck (UGent) and Yvan Saeys (UGent)
- Organization
- Project
-
- Modelling intercellular communication in time and space
- Modelling intercellular signaling events during cellular differentiation from single-cell transcriptomics data
- Unravelling cellular heterogeneity and dynamics in normal and malignant hematopoiesis using singlecell bioinformatics
- Cell-cell cOmmuNicaTion As a driver of Cancer cell state identiTy - Decoding the impact of cell-cell communications on the identity of tumor states in skin cancers
- iPSC-based parenchymal and sinusoidal liver cell mimics for DILI and NAFLD studies (iPSC-LiMic)
- Flanders Artificial Intelligence Research program (FAIR) – second cycle - 2024
- Abstract
- Spatial transcriptomics (ST) technologies allow the profiling of the transcriptome of cells while keeping their spatial context. Since most commercial untargeted ST technologies do not yet operate at single-cell resolution, computational methods such as deconvolution are often used to infer the cell type composition of each sequenced spot. We benchmarked 11 deconvolution methods using 63 silver standards, 3 gold standards, and 2 case studies on liver and melanoma tissues. We developed a simulation engine called synthspot to generate silver standards from single-cell RNA-sequencing data, while gold standards are generated by pooling single cells from targeted ST data. We evaluated methods based on their performance, stability across different reference datasets, and scalability. We found that cell2location and RCTD are the top-performing methods, but surprisingly, a simple regression model outperforms almost half of the dedicated spatial deconvolution methods. Furthermore, we observe that the performance of all methods significantly decreased in datasets with highly abundant or rare cell types. Our results are reproducible in a Nextflow pipeline, which also allows users to generate synthetic data, run deconvolution methods and optionally benchmark them on their dataset (https://github.com/saeyslab/spotless-benchmark).
- Keywords
- spatial transcriptomics, deconvolution, benchmark, single-cell RNA sequencing, Human, Mouse, EXPRESSION, TAXONOMY, SEQ
Downloads
-
5156 24SangAram.pdf
- full text (Published version)
- |
- open access
- |
- |
- 5.11 MB
Citation
Please use this url to cite or link to this publication: http://hdl.handle.net/1854/LU-01J14P7C9RPGVCKXKZDVWER3SG
- MLA
- Sang-aram, Chananchida, et al. “Spotless, a Reproducible Pipeline for Benchmarking Cell Type Deconvolution in Spatial Transcriptomics.” ELIFE, vol. 12, 2024, doi:10.7554/eLife.88431.
- APA
- Sang-aram, C., Browaeys, R., Seurinck, R., & Saeys, Y. (2024). Spotless, a reproducible pipeline for benchmarking cell type deconvolution in spatial transcriptomics. ELIFE, 12. https://doi.org/10.7554/eLife.88431
- Chicago author-date
- Sang-aram, Chananchida, Robin Browaeys, Ruth Seurinck, and Yvan Saeys. 2024. “Spotless, a Reproducible Pipeline for Benchmarking Cell Type Deconvolution in Spatial Transcriptomics.” ELIFE 12. https://doi.org/10.7554/eLife.88431.
- Chicago author-date (all authors)
- Sang-aram, Chananchida, Robin Browaeys, Ruth Seurinck, and Yvan Saeys. 2024. “Spotless, a Reproducible Pipeline for Benchmarking Cell Type Deconvolution in Spatial Transcriptomics.” ELIFE 12. doi:10.7554/eLife.88431.
- Vancouver
- 1.Sang-aram C, Browaeys R, Seurinck R, Saeys Y. Spotless, a reproducible pipeline for benchmarking cell type deconvolution in spatial transcriptomics. ELIFE. 2024;12.
- IEEE
- [1]C. Sang-aram, R. Browaeys, R. Seurinck, and Y. Saeys, “Spotless, a reproducible pipeline for benchmarking cell type deconvolution in spatial transcriptomics,” ELIFE, vol. 12, 2024.
@article{01J14P7C9RPGVCKXKZDVWER3SG,
abstract = {{Spatial transcriptomics (ST) technologies allow the profiling of the transcriptome of cells while keeping their spatial context. Since most commercial untargeted ST technologies do not yet operate at single-cell resolution, computational methods such as deconvolution are often used to infer the cell type composition of each sequenced spot. We benchmarked 11 deconvolution methods using 63 silver standards, 3 gold standards, and 2 case studies on liver and melanoma tissues. We developed a simulation engine called synthspot to generate silver standards from single-cell RNA-sequencing data, while gold standards are generated by pooling single cells from targeted ST data. We evaluated methods based on their performance, stability across different reference datasets, and scalability. We found that cell2location and RCTD are the top-performing methods, but surprisingly, a simple regression model outperforms almost half of the dedicated spatial deconvolution methods. Furthermore, we observe that the performance of all methods significantly decreased in datasets with highly abundant or rare cell types. Our results are reproducible in a Nextflow pipeline, which also allows users to generate synthetic data, run deconvolution methods and optionally benchmark them on their dataset (https://github.com/saeyslab/spotless-benchmark).}},
articleno = {{RP88431}},
author = {{Sang-aram, Chananchida and Browaeys, Robin and Seurinck, Ruth and Saeys, Yvan}},
issn = {{2050-084X}},
journal = {{ELIFE}},
keywords = {{spatial transcriptomics,deconvolution,benchmark,single-cell RNA sequencing,Human,Mouse,EXPRESSION,TAXONOMY,SEQ}},
language = {{eng}},
pages = {{36}},
title = {{Spotless, a reproducible pipeline for benchmarking cell type deconvolution in spatial transcriptomics}},
url = {{http://doi.org/10.7554/eLife.88431}},
volume = {{12}},
year = {{2024}},
}
- Altmetric
- View in Altmetric
- Web of Science
- Times cited: