scMultiGAN: cell-specific imputation for single-cell transcriptomes with multiple deep generative adversarial networks DOI Creative Commons
Tao Wang, Hui Zhao, Yungang Xu

et al.

Briefings in Bioinformatics, Journal Year: 2023, Volume and Issue: 24(6)

Published: Sept. 22, 2023

Abstract The emergence of single-cell RNA sequencing (scRNA-seq) technology has revolutionized the identification cell types and study cellular states at a level. Despite its significant potential, scRNA-seq data analysis is plagued by issue missing values. Many existing imputation methods rely on simplistic distribution assumptions while ignoring intrinsic gene expression specific to cells. This work presents novel deep-learning model, named scMultiGAN, for imputation, which utilizes multiple collaborative generative adversarial networks (GAN). Unlike traditional GAN-based that generate values based random noises, scMultiGAN employs two-stage training process GANs achieve cell-specific imputation. Experimental results show efficacy in accuracy, clustering, differential trajectory analysis, significantly outperforming state-of-the-art techniques. Additionally, scalable large datasets consistently performs well across platforms. code freely available https://github.com/Galaxy8172/scMultiGAN.

Language: Английский

Deciphering spatial domains from spatially resolved transcriptomics with an adaptive graph attention auto-encoder DOI Creative Commons
Kangning Dong, Shihua Zhang

Nature Communications, Journal Year: 2022, Volume and Issue: 13(1)

Published: April 1, 2022

Recent advances in spatially resolved transcriptomics have enabled comprehensive measurements of gene expression patterns while retaining the spatial context tissue microenvironment. Deciphering spots a needs to use their information carefully. To this end, we develop graph attention auto-encoder framework STAGATE accurately identify domains by learning low-dimensional latent embeddings via integrating and profiles. better characterize similarity at boundary domains, adopts an mechanism adaptively learn neighboring spots, optional cell type-aware module through pre-clustering expressions. We validate on diverse datasets generated different platforms with resolutions. could substantially improve identification accuracy denoise data preserving patterns. Importantly, be extended multiple consecutive sections reduce batch effects between extracting three-dimensional (3D) from reconstructed 3D effectively.

Language: Английский

Citations

344
Spatially informed cell-type deconvolution for spatial transcriptomics DOI
Ying Ma, Xiang Zhou

Nature Biotechnology, Journal Year: 2022, Volume and Issue: 40(9), P. 1349 - 1359

Published: May 2, 2022

Language: Английский

Citations

277
Spatially informed clustering, integration, and deconvolution of spatial transcriptomics with GraphST DOI Creative Commons
Yahui Long, Kok Siong Ang, Mengwei Li

et al.

Nature Communications, Journal Year: 2023, Volume and Issue: 14(1)

Published: March 1, 2023

Abstract Spatial transcriptomics technologies generate gene expression profiles with spatial context, requiring spatially informed analysis tools for three key tasks, clustering, multisample integration, and cell-type deconvolution. We present GraphST, a graph self-supervised contrastive learning method that fully exploits data to outperform existing methods. It combines neural networks learn informative discriminative spot representations by minimizing the embedding distance between adjacent spots vice versa. demonstrated GraphST on multiple tissue types technology platforms. achieved 10% higher clustering accuracy better delineated fine-grained structures in brain embryo tissues. is also only can jointly analyze slices vertical or horizontal integration while correcting batch effects. Lastly, superior deconvolution capture niches like lymph node germinal centers exhausted tumor infiltrating T cells breast tissue.

Language: Английский

Citations

233
DeepST: identifying spatial domains in spatial transcriptomics by deep learning DOI Creative Commons

Chang Xu,

Xiyun Jin, Songren Wei

et al.

Nucleic Acids Research, Journal Year: 2022, Volume and Issue: 50(22), P. e131 - e131

Published: Oct. 4, 2022

Recent advances in spatial transcriptomics (ST) have brought unprecedented opportunities to understand tissue organization and function context. However, it is still challenging precisely dissect domains with similar gene expression histology situ. Here, we present DeepST, an accurate universal deep learning framework identify domains, which performs better than the existing state-of-the-art methods on benchmarking datasets of human dorsolateral prefrontal cortex. Further testing a breast cancer ST dataset, showed that DeepST can at finer scale. Moreover, achieve not only effective batch integration data generated from multiple batches or different technologies, but also expandable capabilities for processing other omics data. Together, our results demonstrate has exceptional capacity identifying making desirable tool gain novel insights studies.

Language: Английский

Citations

138
Identifying multicellular spatiotemporal organization of cells with SpaceFlow DOI Creative Commons
Honglei Ren, Benjamin L. Walker, Zixuan Cang

et al.

Nature Communications, Journal Year: 2022, Volume and Issue: 13(1)

Published: July 14, 2022

One major challenge in analyzing spatial transcriptomic datasets is to simultaneously incorporate the cell transcriptome similarity and their locations. Here, we introduce SpaceFlow, which generates spatially-consistent low-dimensional embeddings by incorporating both expression information using spatially regularized deep graph networks. Based on embedding, a pseudo-Spatiotemporal Map that integrates pseudotime concept with locations of cells unravel spatiotemporal patterns cells. By comparing multiple existing methods several at spot single-cell resolutions, SpaceFlow shown produce robust domain segmentation identify biologically meaningful patterns. Applications reveal evolving lineage heart developmental data tumor-immune interactions human breast cancer data. Our study provides flexible learning framework

Language: Английский

Citations

101
Deciphering tissue structure and function using spatial transcriptomics DOI Creative Commons
Benjamin L. Walker, Zixuan Cang, Honglei Ren

et al.

Communications Biology, Journal Year: 2022, Volume and Issue: 5(1)

Published: March 10, 2022

The rapid development of spatial transcriptomics (ST) techniques has allowed the measurement transcriptional levels across many genes together with positions cells. This led to an explosion interest in computational methods and for harnessing both information analysis ST datasets. wide diversity approaches aim, methodology technology provides great challenges dissecting cellular functions contexts. Here, we synthesize review key problems data that are currently applied, while also expanding on open questions areas future development.

Language: Английский

Citations

72
Computational Approaches and Challenges in Spatial Transcriptomics DOI Creative Commons
Shuangsang Fang, Bichao Chen, Yong Zhang

et al.

Genomics Proteomics & Bioinformatics, Journal Year: 2022, Volume and Issue: 21(1), P. 24 - 47

Published: Oct. 14, 2022

The development of spatial transcriptomics (ST) technologies has transformed genetic research from a single-cell data level to two-dimensional coordinate system and facilitated the study composition function various cell subsets in different environments organs. large-scale generated by these ST technologies, which contain gene expression information, have elicited need for spatially resolved approaches meet requirements computational biological interpretation. These include dealing with explosive growth determine cell-level gene-level expression, correcting inner batch effect loss improve quality, conducting efficient interpretation in-depth knowledge mining both at tissue-wide levels, multi-omics integration analysis provide an extensible framework toward understanding processes. However, algorithms designed specifically are still their infancy. Here, we review problems light corresponding issues challenges, present forward-looking insights into algorithm development.

Language: Английский

Citations

72
PROST: quantitative identification of spatially variable genes and domain detection in spatial transcriptomics DOI Creative Commons
Yuchen Liang,

Guowei Shi,

Runlin Cai

et al.

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: Jan. 18, 2024

Abstract Computational methods have been proposed to leverage spatially resolved transcriptomic data, pinpointing genes with spatial expression patterns and delineating tissue domains. However, existing approaches fall short in uniformly quantifying variable (SVGs). Moreover, from a methodological viewpoint, while SVGs are naturally associated depicting domains, they technically dissociated most methods. Here, we present framework (PROST) for the quantitative recognition of patterns, consisting (i) quantitatively characterizing variations gene through PROST Index; (ii) unsupervised clustering domains via self-attention mechanism. We demonstrate that performs superior SVG identification domain segmentation various resolutions, multicellular cellular levels. Importantly, Index can be applied prioritize variations, facilitating exploration biological insights. Together, our study provides flexible robust analyzing diverse data.

Language: Английский

Citations

16
SpaDCN: Deciphering Spatial Functional Landscape from Spatially Resolved Transcriptomics by Aligning Cell–Cell Communications DOI Open Access

Xuefeng Bai,

Xinyu Bao, Chuanchao Zhang

et al.

Small Methods, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 17, 2025

Abstract Spatially resolved transcriptomics (SRT) has emerged as a transformative technology for elucidating cellular organization and tissue architecture. However, significant challenge remains in identifying pathology‐relevant spatial functional landscapes within the microenvironment, primarily due to limited integration of cell–cell communication dynamics. To address this limitation, SpaDCN, Spa tially D ynamic graph C onvolutional N etwork framework is proposed, which aligns communications gene expression context reveal regions with coherent organization. effectively transfer influence on variation, SpaDCN respectively generates node layer edge representation from data ligand–receptor complex contributions then employs dynamic convolution switch propagation graph. It demonstrated that outperforms existing methods domains denoising across various platforms species. Notably, excels marker genes prognostic potential cancer tissues. In conclusion, offers powerful precise tool domain detection transcriptomics, broad applicability types research disciplines.

Language: Английский

Citations

2
Spatial-ID: a cell typing method for spatially resolved transcriptomics via transfer learning and spatial embedding DOI Creative Commons
Rongbo Shen, Lin Liu, Zihan Wu

et al.

Nature Communications, Journal Year: 2022, Volume and Issue: 13(1)

Published: Dec. 10, 2022

Spatially resolved transcriptomics provides the opportunity to investigate gene expression profiles and spatial context of cells in naive state, but at low transcript detection sensitivity or with limited throughput. Comprehensive annotating cell types spatially understand biological processes single level remains challenging. Here we propose Spatial-ID, a supervision-based typing method, that combines existing knowledge reference single-cell RNA-seq data information data. We present series benchmarking analyses on publicly available datasets, demonstrate superiority Spatial-ID compared state-of-the-art methods. Besides, apply self-collected mouse brain hemisphere dataset measured by Stereo-seq, shows scalability three-dimensional large field tissues subcellular resolution.

Language: Английский

Citations

52