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

Topic: Genetics and APOE

Single-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.

Key findings

• Single-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.

New to this subject? Start with our plain-English guide: Genetics →

Study details

• Evidence strengthLow

Read the original

• DOI: 10.1038/s41467-021-22197-x
• PubMed: 33767197
• Original source →

Free copy unknown: no free-copy check has an answer for this record yet. That does not mean it is paywalled.

@article{wang2021-33767197,
  author  = {Wang, J. and Ma, A. and Chang, Y. and Gong, J. and Jiang, Y. and Qi, R. and Wang, C. and Fu, H. and others},
  title   = {{scGNN is a novel graph neural network framework for single-cell RNA-Seq analyses}},
  journal = {Nature communications},
  year    = {2021},
  doi     = {10.1038/s41467-021-22197-x},
  pmid    = {33767197},
  url     = {https://doi.org/10.1038/s41467-021-22197-x}
}

Related studies in Genetics

• A single-cell and spatial RNA-seq database for Alzheimer's disease (ssREAD) Nature communications · 2024-06-06 · usefulness score 46/100
• Large-Scale Integration of Single-Cell RNA-Seq Data Reveals Astrocyte Diversity and Transcriptomic Modules across Six Central Nervous System Disorders Biomolecules · 2023-04-19 · usefulness score 41/100
• scREAD: A Single-Cell RNA-Seq Database for Alzheimer's Disease iScience · 2020-11-05 · usefulness score 39/100
• Conditional polygenic enrichment distinguishes causal from tagging disease-critical cell populations in single-cell RNA-seq medRxiv (preprint) · 2026-08-23 · usefulness score 16/100
• Single-Nucleus RNA-Seq Is Not Suitable for Detection of Microglial Activation Genes in Humans Cell reports · 2020-09-01 · usefulness score 41/100
• A single-nuclei multiomics resource across four brain regions prioritises human neural cell types influencing brain-related traits bioRxiv · 2026-09-25 · usefulness score 16/100

Also tagged “epigenetic”

274 studies in the library carry the topic tag “epigenetic”; these are this study's neighbours in a fixed, alphabetical-by-hash tour of them, not the closest in content. All 274 tagged studies.

• Longitudinal single-cell transcriptional profiling identifies dynamic neurodegenerative mechanisms in Christianson Syndrome rat bioRxiv · 2026-10-20 · usefulness score 16/100 · Risk genes beyond APOE
• Circular Data Analysis for Spatial Omics bioRxiv · 2026-10-06 · usefulness score 16/100 · Risk factors
• Reprogramming neurodegeneration in the big data era Current opinion in neurobiology · 2018-01-10 · usefulness score 39/100 · Genetics

Similar studies

• hdWGCNA identifies co-expression networks in high-dimensional transcriptomics data 2023 · Genetics
• Guidelines for bioinformatics of single-cell sequencing data analysis in Alzheimer's disease: review, recommendation, implementation and application 2022 · Genetics
• Single-cell transcriptomics unveils gene regulatory network plasticity 2019 · Genetics
• Identifying phenotype-associated subpopulations by integrating bulk and single-cell sequencing data 2022 · Genetics
• Phenotype Scoring of Population Scale Single-Cell Data Dissects Alzheimer's Disease Complexity 2024 · Genetics
• SCAN: Spatiotemporal Cloud Atlas for Neural cells 2024 · Genetics

More on Genetics

• A review that covers this ground Identification of 16 novel Alzheimer's disease loci using multi-ancestry meta-analyses Alzheimer's & Dementia · 2025-02 · usefulness score 66/100
• Newest in this subject APH-1A and CDK5 Expression Correlates with Tau Hyperphosphorylation and Its Attenuation by Brahmi and Donep-M in STZ induced Alzheimer’s Disease Model Research Square · 2026-10-31 · Preprint — not yet peer-reviewed · usefulness score 16/100
• Newest in this subject Integrative systems biology identifies shared neuroimmune co-expression programmes across pulmonary tuberculosis and Alzheimer's disease Research Square · 2026-10-30 · Preprint — not yet peer-reviewed · usefulness score 16/100

See all studies in Genetics →