Adversarial domain translation networks for integrating large-scale atlas-level single-cell datasets

Zhao, Jidong; Wang, Gefei; Ming, Jingsi; Lin, Zhixiang; Wang, Yan; Agarwal, Snigdha; Agrawal, Aditi; Al-Moujahed, Ahmad; Alam, Alina; Albertelli, Megan A.; Allegakoen, Paul; Ambrosi, Thomas H.; Antony, Jane; Artandi, Steven E.; Aujard, Fabienne; Awayan, Kyle; Baghel, Ankit S.; Bakerman, Isaac; Bakken, Trygve E.; Baruni, Jalal; Beachy, Philip A.; Bilen, Biter; Botvinnik, Olga; Boyd, Scott D.; Burhan, Deviana; Casey, Kerriann M.; Chan, Charles; Chang, Charles; Chang, Stephen; Ming, Chen; Clarke, Michael F.; Crasta, Sheela; Culver, Rebecca; D’Addabbo, Jessica; Darmanis, Spyros; Dehghannasiri, Roozbeh; Ding, Shilei; Duffy, Connor V.; Epelbaum, Jacques; Espinoza, Francisco; Ezran, Camille; Farup, Jean; Ferrell, James E.; Frank, Hannah K.; Fuller, Margaret T.; Gillich, Astrid; Godoy, Elias; Gratzinger, Dita; Guethlein, Lisbeth A.; Hang, Yan; Hasegawa, Kazuteru; Hodge, Rebecca D.; Hoover, Malachia; Huang, Franklin W.; Huang, Kerwyn Casey; Huynh, Shelly; Isobe, Taichi; Israel, Carly; Jang, SoRi; Jing, Qiuyu; Jones, Robert C.; Kang, Jengmin; Karanewsky, Caitlin J.; Karkanias, Jim; Kebschull, Justus M.; Kershner, Aaron M.; Kim, Lily; Kim, Seung K.; Kirk, E. Christopher; Koh, Winston; Konermann, Silvana; Kong, William; Krasnow, Mark A.; Kuo, Christin S.; Lautier, Corinne; Lee, Song Eun; Lein, Ed; Lewis, Rebecca; Li, Peng; Lin, Shengda; Liu, Shixuan; Liu, Yin; Loeb, Gabriel B.; Long, Jonathan Z.; Lu, Wan-Jin; Lucot, Katherine L; Luo, Liqun; McGeever, Aaron; Metzger, Ross J.; Montine, Tom; Morrée, Antoine de; Morri, Maurizio; Mrouj, Karim; Mukherjee, Shravani; Nabhan, Ahmad; Nafees, Saba; Neff, Norma; Neuhöfer, Patrick; Nguyen, Patricia K.; Okamoto, Jennifer; Olivieri, Julia Eve; Ouadah, Youcef; Paine, Honor; Parham, Peter; Pendleton, Jozeph L.; Penland, Lolita; Perret, Martine; Pisco, Angela Oliveira; Qi, Zhengjian; Quake, Stephen R.; Radespiel, Ute; Rando, Thomas A.; Ravelonjanahary, Hajanirina Noëline; Razafindrakoto, Andriamahery; Salzman, Julia; Schaum, Nicholas; Schopler, Robert; Scott, Bronwyn; Shapiro, Liza J.; Sin, Ho-Su; Sinha, Rahul; Sit, Rene; Stanley, Geoff; Stryer, Lubert; Subramaniam, Varun Ramanan; Swarup, Aditi; Tan, Weilun; Tarashansky, Alexander J.; Taychameekiatchai, Aris; Terrien, Jérémy; Travaglini, Kyle J.; Urtasun, Andoni; Sivakamasundari,; Veerakumar, Avin; Vemuri, Venkata Naga Pranathi; Verdier, Jean‐Michel; Vlaminck, Iwijn De; Vollrath, Douglas; Wang, Bo; Wang, Bruce; Wang, Michael F. Z.; Wang, Sheng; Webber, James T.; Weinstein, Hannah; Weissman, Irving L.; Wiggenhorn, Amanda L.; Williams, Cathy V.; Wright, Patricia C.; Wu, Albert Y.; Wu, Angela Ruohao; Wyss‐Coray, Tony; Bao, Xiang; Jia, Yan; Yang, Can; Yang, Ji-Qin; Yoder, Anne D.; Yu, Brian; Yung, Andrea R.; Zhang, Yue; Zhao, Zhuofeng

doi:10.1038/s43588-022-00251-y

Cited by 21 publications

(3 citation statements)

References 71 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, we added Harmony 37 , which was demonstrated to have the best performance and shortest running time in previous batch-effect removal benchmark study in scRNA-seq data 38 . In the meantime, we also included Portal 39 , a recently published integration method which has not been benchmarked and can take gene scores as input. After performing the integration between reference and target datasets, we apply MLP as classifier to transfer cell labels according to the integrated output (details in Methods section).…”

Section: Resultsmentioning

confidence: 99%

Cellcano: supervised cell type identification for single cell ATAC-seq data

2023

Nat Commun

View full text Add to dashboard Cite

Computational cell type identification is a fundamental step in single-cell omics data analysis. Supervised celltyping methods have gained increasing popularity in single-cell RNA-seq data because of the superior performance and the availability of high-quality reference datasets. Recent technological advances in profiling chromatin accessibility at single-cell resolution (scATAC-seq) have brought new insights to the understanding of epigenetic heterogeneity. With continuous accumulation of scATAC-seq datasets, supervised celltyping method specifically designed for scATAC-seq is in urgent need. Here we develop Cellcano, a computational method based on a two-round supervised learning algorithm to identify cell types from scATAC-seq data. The method alleviates the distributional shift between reference and target data and improves the prediction performance. After systematically benchmarking Cellcano on 50 well-designed celltyping tasks from various datasets, we show that Cellcano is accurate, robust, and computationally efficient. Cellcano is well-documented and freely available at https://marvinquiet.github.io/Cellcano/.

show abstract

Section: Resultsmentioning

confidence: 99%

Cellcano: supervised cell type identification for single cell ATAC-seq data

2023

Nat Commun

View full text Add to dashboard Cite

show abstract

“…By effectively designing the shared latent space and accounting for slice-specific and gene-specific effects, STitch3D is able to preserve biologically meaningful information that varies along the z-axis, while also removing batch effects across slices. Additionally, it has been shown by previous studies that integrating more data improves statistical power in single-cell data analysis [40,41,42,43]. Here, as a large amount of information is shared across slices, such as similar gene expression levels in adjacent slices, STitch3D's joint analysis of multiple slices also improves accuracy.…”

Section: Discussionmentioning

confidence: 98%

Construction of a 3D whole organism spatial atlas by joint modeling of multiple slices

Wang

Zhao

Yan

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Spatial transcriptomics (ST) technologies are revolutionizing the way that researchers explore the spatial architecture of tissues. Currently, ST data analysis is often restricted to 2D space within a single tissue slice, limiting our capacity to understand biological processes that take place in 3D space. Here, we present STitch3D, a unified computational framework that integrates multiple 2D tissue slices to reconstruct 3D cellular structures from the tissue level to the whole organism level. By jointly modeling multiple 2D tissue slices and integrating them with cell-type-specific expression profiles derived from single-cell RNA-sequencing data, STitch3D simultaneously identifies 3D spatial regions with coherent gene expression levels and reveals 3D distributions of cell types. STitch3D distinguishes biological variation among slices from batch effects, and effectively borrows shared information across slices to assemble powerful 3D models of tissues. Through comprehensive experiments using diverse datasets, we demonstrate the performance of STitch3D in building comprehensive 3D tissue architectures of the mouse brain, the human heart, and the Drosophila embryo, which allow 3D analysis in the entire tissue region or even the whole organism. To gain deeper biological insights, the outputs of STitch3D can be used for downstream tasks, such as inference of spatial trajectories, identification of spatially variable genes enriched in tissue regions or subregions, denoising or imputation of spatial gene expressions, as well as generation of virtual tissue slices.

show abstract

“…Cells that were deemed low quality (i.e., low in transcript and/or gene count) and technical doublets were removed from downstream analysis. Next, to ensure cell types are comparable across species, we applied Portal 135 to integrate data (one-to-one orthologs only) from different species. Portal projects data into a space that minimizes species differences, from which an integrated UMAP is generated to visualize cell clustering from different species.…”

Section: Methodsmentioning

confidence: 99%

An organism-wide atlas of hormonal signaling based on the mouse lemur single-cell transcriptome

Liu,

Ezran,

Wang

et al. 2024

Nat Commun

Self Cite

View full text Add to dashboard Cite

Hormones mediate long-range cell communication and play vital roles in physiology, metabolism, and health. Traditionally, endocrinologists have focused on one hormone or organ system at a time. Yet, hormone signaling by its very nature connects cells of different organs and involves crosstalk of different hormones. Here, we leverage the organism-wide single cell transcriptional atlas of a non-human primate, the mouse lemur (Microcebus murinus), to systematically map source and target cells for 84 classes of hormones. This work uncovers previously-uncharacterized sites of hormone regulation, and shows that the hormonal signaling network is densely connected, decentralized, and rich in feedback loops. Evolutionary comparisons of hormonal genes and their expression patterns show that mouse lemur better models human hormonal signaling than mouse, at both the genomic and transcriptomic levels, and reveal primate-specific rewiring of hormone-producing/target cells. This work complements the scale and resolution of classical endocrine studies and sheds light on primate hormone regulation.

show abstract

Adversarial domain translation networks for integrating large-scale atlas-level single-cell datasets

Cited by 21 publications

References 71 publications

Cellcano: supervised cell type identification for single cell ATAC-seq data

Cellcano: supervised cell type identification for single cell ATAC-seq data

Construction of a 3D whole organism spatial atlas by joint modeling of multiple slices

An organism-wide atlas of hormonal signaling based on the mouse lemur single-cell transcriptome

Contact Info

Product

Resources

About