scholarly journals scGNN: a novel graph neural network framework for single-cell RNA-Seq analyses

2020 ◽  
Author(s):  
Juexin Wang ◽  
Anjun Ma ◽  
Yuzhou Chang ◽  
Jianting Gong ◽  
Yuexu Jiang ◽  
...  
Keyword(s):  
Neural Network ◽  
Gene Expression ◽  
Single Cell ◽  
Neural Development ◽  
Expression Patterns ◽  
Gaussian Model ◽  
Postmortem Brain ◽  
Differential Mechanism ◽  
Disease Study ◽  
Cell Cell

ABSTRACTSingle-cell RNA-sequencing (scRNA-Seq) is widely used to reveal the heterogeneity and dynamics of tissues, organisms, and complex diseases, but its analyses still suffer from multiple grand challenges, including the sequencing sparsity and complex differential patterns in gene expression. We introduce the scGNN (single-cell graph neural network) to provide a hypothesis-free deep learning framework for scRNA-Seq analyses. This framework formulates and aggregates cell-cell relationships with graph neural networks and models heterogeneous gene expression patterns using a left-truncated mixture Gaussian model. scGNN integrates three iterative multi-modal autoencoders and outperforms existing tools for gene imputation and cell clustering on four benchmark scRNA-Seq datasets. In an Alzheimer’s disease study with 13,214 single nuclei from postmortem brain tissues, scGNN successfully illustrated disease-related neural development and the differential mechanism. scGNN provides an effective representation of gene expression and cell-cell relationships. It is also a novel and powerful framework that can be applied to scRNA-Seq analyses.

scholarly journals scGNN is a novel graph neural network framework for single-cell RNA-Seq analyses

2021 ◽  
Vol 12 (1) ◽  
Author(s):  
Juexin Wang ◽  
Anjun Ma ◽  
Yuzhou Chang ◽  
Jianting Gong ◽  
Yuexu Jiang ◽  
...  
Keyword(s):  
Neural Network ◽  
Gene Expression ◽  
Single Cell ◽  
Neural Development ◽  
Expression Patterns ◽  
Gaussian Model ◽  
Postmortem Brain ◽  
Differential Mechanism ◽  
Disease Study ◽  
Cell Cell

AbstractSingle-cell RNA-sequencing (scRNA-Seq) is widely used to reveal the heterogeneity and dynamics of tissues, organisms, and complex diseases, but its analyses still suffer from multiple grand challenges, including the sequencing sparsity and complex differential patterns in gene expression. We introduce the scGNN (single-cell graph neural network) to provide a hypothesis-free deep learning framework for scRNA-Seq analyses. This framework formulates and aggregates cell–cell relationships with graph neural networks and models heterogeneous gene expression patterns using a left-truncated mixture Gaussian model. scGNN integrates three iterative multi-modal autoencoders and outperforms existing tools for gene imputation and cell clustering on four benchmark scRNA-Seq datasets. In an Alzheimer’s disease study with 13,214 single nuclei from postmortem brain tissues, scGNN successfully illustrated disease-related neural development and the differential mechanism. scGNN provides an effective representation of gene expression and cell–cell relationships. It is also a powerful framework that can be applied to general scRNA-Seq analyses.

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

2021 ◽  
Author(s):  
Shixuan Liu ◽  
Camille Ezran ◽  
Michael F.Z. Wang ◽  
Zhengda Li ◽  
Jonathan Z. Long ◽  
...  
Keyword(s):  
Gene Expression ◽  
Single Cell ◽  
Expression Patterns ◽  
Cell Communication ◽  
Cell Types ◽  
Mouse Lemur ◽  
Target Cells ◽  
Endocrine Gland ◽  
Hormone Regulation ◽  
Cell Cell

Hormones coordinate long-range cell-cell communication in multicellular organisms and play vital roles in normal physiology, metabolism, and health. Using the newly-completed organism-wide single cell transcriptional atlas of a non-human primate, the mouse lemur (Microcebus murinus), we have systematically identified hormone-producing and -target cells for 87 classes of hormones, and have created a browsable atlas for hormone signaling that reveals previously unreported sites of hormone regulation and species-specific rewiring. Hormone ligands and receptors exhibited cell-type-dependent, stereotypical expression patterns, and their transcriptional profiles faithfully classified the discrete cell types defined by the full transcriptome, despite their comprising less than 1% of the transcriptome. Although individual cell types generally exhibited the same characteristic patterns of hormonal gene expression, a number of examples of similar or seemingly-identical cell types (e.g., endothelial cells of the lung versus of other organs) displayed different hormonal gene expression patterns. By linking ligand-expressing cells to the cells expressing the corresponding receptor, we constructed an organism-wide map of the hormonal cell-cell communication network. The hormonal cell-cell network was remarkably densely and robustly connected, and included classical hierarchical circuits (e.g. pituitary → peripheral endocrine gland → diverse cell types) as well as examples of highly distributed control. The network also included both well-known examples of feedback loops and a long list of potential novel feedback circuits. This primate hormone atlas provides a powerful resource to facilitate discovery of regulation on an organism-wide scale and at single-cell resolution, complementing the single-site-focused strategy of classical endocrine studies. The network nature of hormone regulation and the principles discovered here further emphasize the importance of a systems approach to understanding hormone regulation.

scholarly journals Identifying cell types from single-cell data based on similarities and dissimilarities between cells

2021 ◽  
Vol 22 (S3) ◽  
Author(s):  
Yuanyuan Li ◽  
Ping Luo ◽  
Yi Lu ◽  
Fang-Xiang Wu
Keyword(s):  
Gene Expression ◽  
Single Cell ◽  
Spectral Clustering ◽  
Incidence Matrix ◽  
Expression Patterns ◽  
Cell Types ◽  
Clustering Method ◽  
Different Types ◽  
Cell Data ◽  
Spectral Clustering Method

Abstract Background With the development of the technology of single-cell sequence, revealing homogeneity and heterogeneity between cells has become a new area of computational systems biology research. However, the clustering of cell types becomes more complex with the mutual penetration between different types of cells and the instability of gene expression. One way of overcoming this problem is to group similar, related single cells together by the means of various clustering analysis methods. Although some methods such as spectral clustering can do well in the identification of cell types, they only consider the similarities between cells and ignore the influence of dissimilarities on clustering results. This methodology may limit the performance of most of the conventional clustering algorithms for the identification of clusters, it needs to develop special methods for high-dimensional sparse categorical data. Results Inspired by the phenomenon that same type cells have similar gene expression patterns, but different types of cells evoke dissimilar gene expression patterns, we improve the existing spectral clustering method for clustering single-cell data that is based on both similarities and dissimilarities between cells. The method first measures the similarity/dissimilarity among cells, then constructs the incidence matrix by fusing similarity matrix with dissimilarity matrix, and, finally, uses the eigenvalues of the incidence matrix to perform dimensionality reduction and employs the K-means algorithm in the low dimensional space to achieve clustering. The proposed improved spectral clustering method is compared with the conventional spectral clustering method in recognizing cell types on several real single-cell RNA-seq datasets. Conclusions In summary, we show that adding intercellular dissimilarity can effectively improve accuracy and achieve robustness and that improved spectral clustering method outperforms the traditional spectral clustering method in grouping cells.

scholarly journals Cryopreservation of microglia enables single-cell RNA sequencing with minimal effects on disease-related gene expression patterns

iScience ◽  
2021 ◽  
Vol 24 (4) ◽  
pp. 102357
Author(s):  
Brenda Morsey ◽  
Meng Niu ◽  
Shetty Ravi Dyavar ◽  
Courtney V. Fletcher ◽  
Benjamin G. Lamberty ◽  
...  
Keyword(s):  
Gene Expression ◽  
Single Cell ◽  
Rna Sequencing ◽  
Expression Patterns ◽  
Gene Expression Patterns ◽  
Related Gene ◽  
Single Cell Rna Sequencing ◽  
Disease Related Gene

scholarly journals Joint profiling of gene expression and chromatin accessibility of amphioxus development at single cell resolution

2021 ◽  
Author(s):  
Pengcheng Ma ◽  
Xingyan Liu ◽  
Huimin Liu ◽  
Zaoxu Xu ◽  
Xiangning Ding ◽  
...  
Keyword(s):  
Gene Expression ◽  
Single Cell ◽  
Cell Fate ◽  
Regulatory Networks ◽  
Functional Divergence ◽  
Expression Patterns ◽  
Genetic Regulatory Networks ◽  
Public Access ◽  
Origin And Evolution ◽  
Key Species

Abstract Vertebrate evolution was accompanied with two rounds of whole genome duplication followed by functional divergence in terms of regulatory circuits and gene expression patterns. As a basal and slow-evolving chordate species, amphioxus is an ideal paradigm for exploring the origin and evolution of vertebrates. Single cell sequencing has been widely employed to construct the developmental cell atlas of several key species of vertebrates (human, mouse, zebrafish and frog) and tunicate (sea squirts). Here, we performed single-nucleus RNA sequencing (snRNA-seq) and single-cell assay for transposase accessible chromatin sequencing (scATAC-seq) for different stages of amphioxus (covering embryogenesis and adult tissues). With the datasets generated we constructed the developmental tree for amphioxus cell fate commitment and lineage specification, and revealed the underlying key regulators and genetic regulatory networks. The generated data were integrated into an online platform, AmphioxusAtlas, for public access at http://120.79.46.200:81/AmphioxusAtlas.

scholarly journals ZipSeq : Barcoding for Real-time Mapping of Single Cell Transcriptomes

2020 ◽  
Author(s):  
Kenneth H. Hu ◽  
John P. Eichorst ◽  
Chris S. McGinnis ◽  
David M. Patterson ◽  
Eric D. Chow ◽  
...  
Keyword(s):  
Gene Expression ◽  
Single Cell ◽  
Real Time ◽  
Dimensional Space ◽  
Single Cells ◽  
Expression Patterns ◽  
Live Cells ◽  
Glass Substrates ◽  
Complete Mapping

ABSTRACTSpatial transcriptomics seeks to integrate single-cell transcriptomic data within the 3-dimensional space of multicellular biology. Current methods use glass substrates pre-seeded with matrices of barcodes or fluorescence hybridization of a limited number of probes. We developed an alternative approach, called ‘ZipSeq’, that uses patterned illumination and photocaged oligonucleotides to serially print barcodes (Zipcodes) onto live cells within intact tissues, in real-time and with on-the-fly selection of patterns. Using ZipSeq, we mapped gene expression in three settings: in-vitro wound healing, live lymph node sections and in a live tumor microenvironment (TME). In all cases, we discovered new gene expression patterns associated with histological structures. In the TME, this demonstrated a trajectory of myeloid and T cell differentiation, from periphery inward. A variation of ZipSeq efficiently scales to the level of single cells, providing a pathway for complete mapping of live tissues, subsequent to real-time imaging or perturbation.

scholarly journals 1450-1545 Young Investigator Awards & Presentations Basic/Translational Exploring cellular subpopulations in glioblastoma and matched organoids using single-cell RNA-seq 52

2018 ◽  
Vol 45 (S3) ◽  
pp. S13-S14
Author(s):  
VG LeBlanc ◽  
D Trinh ◽  
M Hughes ◽  
I Luthra ◽  
D Livingstone ◽  
...  
Keyword(s):  
Gene Expression ◽  
Single Cell ◽  
Tissue Sample ◽  
Expression Patterns ◽  
Model Systems ◽  
Sequencing Data ◽  
Tissue Samples ◽  
Original Tumour ◽  
Current Therapies ◽  
Intratumoural Heterogeneity

Glioblastomas (GBMs) account for nearly half of all primary malignant brain tumours, and current therapies are often only marginally effective. Our understanding of the underlying biology of these tumours and the development of new therapies have been complicated in part by widespread inter- and intratumoural heterogeneity. To characterize this heterogeneity, we performed regional subsampling of primary glioblastomas and derived organoids from these tissue samples. We then performed single-cell RNA-sequencing (scRNA-seq) on these primary regional subsamples and 1-3 matched organoids per sample. We have profiled samples from six tumour sets to date and have obtained sequencing data for 21,234 primary tissue cells and 14,742 organoid cells. While the most apparent differences in gene expression appear to be between individual tumours, we were also able to identify similar cellular subpopulations across tissue samples and across organoids. Importantly, organoids derived from the same tissue sample appeared to be composed of similar cellular subpopulations and were highly comparable to each other, indicating that replicate organoids faithfully represent the original tumour tissue. Overall, our scRNA-seq approach will help evaluate the utility of tumour-derived organoids as model systems for GBM and will aid in identifying cellular subpopulations defined by gene expression patterns, both in primary GBM regional subsamples and their associated organoids. These analyses will allow for the characterization of clonal or subclonal populations that are likely to respond to different therapeutic approaches and may also uncover novel therapeutic targets previously unrevealed through bulk analyses.

Cancer classification of single-cell gene expression data by neural network

2019 ◽  
Author(s):  
Bong-Hyun Kim ◽  
Kijin Yu ◽  
Peter C W Lee
Keyword(s):  
Neural Network ◽  
Gene Expression ◽  
Single Cell ◽  
Expression Profiles ◽  
Gene Expression Profiles ◽  
Cancer Classification ◽  
Supplementary Information ◽  
Support Vector ◽  
K Nearest Neighbors ◽  
Normal Tissues

Abstract Motivation Cancer classification based on gene expression profiles has provided insight on the causes of cancer and cancer treatment. Recently, machine learning-based approaches have been attempted in downstream cancer analysis to address the large differences in gene expression values, as determined by single-cell RNA sequencing (scRNA-seq). Results We designed cancer classifiers that can identify 21 types of cancers and normal tissues based on bulk RNA-seq as well as scRNA-seq data. Training was performed with 7398 cancer samples and 640 normal samples from 21 tumors and normal tissues in TCGA based on the 300 most significant genes expressed in each cancer. Then, we compared neural network (NN), support vector machine (SVM), k-nearest neighbors (kNN) and random forest (RF) methods. The NN performed consistently better than other methods. We further applied our approach to scRNA-seq transformed by kNN smoothing and found that our model successfully classified cancer types and normal samples. Availability and implementation Cancer classification by neural network. Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals PRIME: a probabilistic imputation method to reduce dropout effects in single-cell RNA sequencing

2020 ◽  
Vol 36 (13) ◽  
pp. 4021-4029
Author(s):  
Hyundoo Jeong ◽  
Zhandong Liu
Keyword(s):  
Gene Expression ◽  
Single Cell ◽  
Rna Sequencing ◽  
Expression Profiles ◽  
Expression Patterns ◽  
Imputation Method ◽  
Supplementary Information ◽  
Single Cell Sequencing ◽  
Depth Analysis ◽  
Single Cell Rna Sequencing

Abstract Summary Single-cell RNA sequencing technology provides a novel means to analyze the transcriptomic profiles of individual cells. The technique is vulnerable, however, to a type of noise called dropout effects, which lead to zero-inflated distributions in the transcriptome profile and reduce the reliability of the results. Single-cell RNA sequencing data, therefore, need to be carefully processed before in-depth analysis. Here, we describe a novel imputation method that reduces dropout effects in single-cell sequencing. We construct a cell correspondence network and adjust gene expression estimates based on transcriptome profiles for the local subnetwork of cells of the same type. We comprehensively evaluated this method, called PRIME (PRobabilistic IMputation to reduce dropout effects in Expression profiles of single-cell sequencing), on synthetic and eight real single-cell sequencing datasets and verified that it improves the quality of visualization and accuracy of clustering analysis and can discover gene expression patterns hidden by noise. Availability and implementation The source code for the proposed method is freely available at https://github.com/hyundoo/PRIME. Supplementary information Supplementary data are available at Bioinformatics online.

scholarly journals Microfluidic single-cell real-time PCR for comparative analysis of gene expression patterns

2012 ◽  
Vol 7 (5) ◽  
pp. 829-838 ◽  
Author(s):  
Veronica Sanchez-Freire ◽  
Antje D Ebert ◽  
Tomer Kalisky ◽  
Stephen R Quake ◽  
Joseph C Wu
Keyword(s):  
Gene Expression ◽  
Comparative Analysis ◽  
Single Cell ◽  
Real Time ◽  
Real Time Pcr ◽  
Expression Patterns ◽  
Gene Expression Patterns
Sign in / Sign up

Export Citation Format

Share Document