CellDART: Cell type inference by domain adaptation of single-cell and spatial transcriptomic data

Bae, Sungwoo; Na, Kwon Joong; Koh, Jaemoon; Lee, Dong Hoon; Choi, Hongyoon; Kim, Young Tae

doi:10.1101/2021.04.26.441459

Cited by 8 publications

(19 citation statements)

References 53 publications

(70 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Cell2Location utilizes a Bayesian model to predict cell type identities ( Kleshchevnikov et al, 2020 ). Lastly, CellDART tackles the problem of identifying the cell type composition of each spot by leveraging neural networks ( Bae et al, 2021 ).…”

Section: Decomposing Spatial Transcriptomic Spots Into Multiple Cell Typesmentioning

confidence: 99%

Principles of Spatial Transcriptomics Analysis: A Practical Walk-Through in Kidney Tissue

et al. 2022

View full text Add to dashboard Cite

Spatial transcriptomic technologies capture genome-wide readouts across biological tissue space. Moreover, recent advances in this technology, including Slide-seqV2, have achieved spatial transcriptomic data collection at a near-single cell resolution. To-date, a repertoire of computational tools has been developed to discern cell type classes given the transcriptomic profiles of tissue coordinates. Upon applying these tools, we can explore the spatial patterns of distinct cell types and characterize how genes are spatially expressed within different cell type contexts. The kidney is one organ whose function relies upon spatially defined structures consisting of distinct cellular makeup. Thus, the application of Slide-seqV2 to kidney tissue has enabled us to elucidate spatially characteristic cellular and genetic profiles at a scale that remains largely unexplored. Here, we review spatial transcriptomic technologies, as well as computational approaches for cell type mapping and spatial cell type and transcriptomic characterizations. We take kidney tissue as an example to demonstrate how the technologies are applied, while considering the nuances of this architecturally complex tissue.

show abstract

Section: Decomposing Spatial Transcriptomic Spots Into Multiple Cell Typesmentioning

confidence: 99%

Principles of Spatial Transcriptomics Analysis: A Practical Walk-Through in Kidney Tissue

et al. 2022

View full text Add to dashboard Cite

show abstract

“…We then analyzed the cell distribution of the clusters using three different types of cell-type prediction methods, in this case multimodal intersection analysis (MIA) 25 , robust cell type decomposition (RCTD) 26 , and single-cell and spatial transcriptomic data (CellDART) 27 . Using MIA, fibroblasts and endothelial cells were preferentially discovered in cluster 1 while cancer cells were found to be predominant in cluster 2 ( Figure 4B ).…”

Section: Resultsmentioning

confidence: 99%

“…All the parameters were set to the default settings including doublet_mode of ‘ doublet ’ in run.RCTD ( Supplementary Material 8 ). Another algorithm, CellDART 27 , which used adversarial domain adaptation classification from single-cell data with pre-labeled cell types was additionally performed to find cell types related to the distribution of fluorescence ( Supplementary Material 9 ).…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Spatial Transcriptomics-based Identification of Molecular Markers for Nanomedicine Distribution in Tumor Tissue

Park

Choi

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

The intratumoral accumulation of nanomedicine has been considered a passive process, referred to as the enhanced permeability and retention (EPR) effect. Recent studies have suggested that the tumor uptake of nanomedicines follows an energy-dependent pathway rather than being a passive process. Herein, to explore the factor candidates that are associated with nanomedicine tumor uptake, we developed a molecular marker identification platform by integrating microscopic fluorescence images of a nanomedicine distribution with spatial transcriptomics (ST) information. When this approach is applied to PEGylated liposomes, molecular markers related to hypoxia, glucose metabolism and apoptosis can be identified as being related to the intratumoral distribution of the nanomedicine. We expect that our method can be applied to explain the distribution of a wide range of nanomedicines and that the data obtained from this analysis can enhance the precise utilization of nanomedicines.

show abstract

“…One of the limitations of spatial transcriptomics data is the fact that spots are not directly interpreted as cells. Therefore, multiple computational approaches have been suggested for accurate spatial mapping of cell types by integrating spatial and single-cell transcriptomics [3][4][5][6][7][8][9][10] . They can be further utilized to speculate the spatial infiltration pattern of a few cell types that play a key role in the pathophysiology of various diseases [11][12][13][14] .…”

Section: Mainmentioning

confidence: 99%

spSeudoMap: Cell type mapping of spatial transcriptomics using unmatched single-cell RNA-seq data

Bae

Choi

Lee

2022

Preprint

Self Cite

View full text Add to dashboard Cite

With advances in computational models, the cellular landscape can be tracked in various tissues using spatial transcriptomics. Since many single-cell RNA-seq (scRNA-seq) data have been obtained after cell sorting, such as when investigating immune cells, integrating these singlecell data with spatial data is limited due to a mismatch of cell types composing the two datasets. Here, we present a method, spSeudoMap, which utilizes sorted scRNA-seq data to train a model for predicting cell types of spatial spots by creating virtual cell mixtures that closely mimic the gene expression profile of spatial transcriptomic data. To overcome the mismatch issue, the cell type exclusively present in the spatial data, pseudotype, was defined. The proportion of pseudotype cells and virtual expression profiles in the cell mixture was determined by pseudobulk transcriptomes. The simulated cell mixture was considered a reference dataset, and the model that predicts the cell composition of the mixture was trained to predict the cell fraction of the spatial data using domain adaptation. First, spSeudoMap was evaluated in human and mouse brain tissues, and the main region-specific neuron types extracted from single-cell data could be precisely mapped to the expected anatomical locations. Moreover, the method was applied to human breast cancer data and described the spatial distribution of immune cell subtypes and their interactions in heterogeneous tissue. Taken together, spSeudoMap is a platform that predicts the spatial composition of cell subpopulations using sorted scRNA-seq data, and it may help to clarify the roles of a few but crucial cell types.

show abstract

CellDART: Cell type inference by domain adaptation of single-cell and spatial transcriptomic data

Cited by 8 publications

References 53 publications

Principles of Spatial Transcriptomics Analysis: A Practical Walk-Through in Kidney Tissue

Principles of Spatial Transcriptomics Analysis: A Practical Walk-Through in Kidney Tissue

Spatial Transcriptomics-based Identification of Molecular Markers for Nanomedicine Distribution in Tumor Tissue

spSeudoMap: Cell type mapping of spatial transcriptomics using unmatched single-cell RNA-seq data

Contact Info

Product

Resources

About