Aug 2026· International Journal of Molecular Sciences· Vol 27· 0 citations· 48 references
Medicine
TL;DR
A structure-aware interleaved-attention graph learning framework, termed IAGRN, is proposed for GRN inference from scRNA-seq data that interleaves topology-constrained local attention with distance-aware global attention, enabling effective integration of structural priors and long-range regulatory signals.
Abstract
Gene regulatory networks (GRNs) describe regulatory interactions between transcription factors and their target genes and are essential for understanding cellular processes and disease mechanisms. Recent advances in single-cell RNA sequencing (scRNA-seq) have enabled data-driven GRN inference at single-cell resolution. However, the high sparsity and noise inherent in scRNA-seq data pose substantial challenges for accurately recovering regulatory relationships. Existing graph neural network (GNN)-based approaches often rely on localized message passing, which can lead to over-smoothing and limited modeling of long-range regulatory dependencies. To address these limitations, a structure-aware interleaved-attention graph learning framework, termed IAGRN, is proposed for GRN inference from scRNA-seq data. Specifically, it interleaves topology-constrained local attention with distance-aware global attention, enabling effective integration of structural priors and long-range regulatory signals. Graph Laplacian positional encoding is further incorporated to preserve topological information and enhance node representations. Evaluations on seven public benchmark datasets demonstrate that IAGRN consistently improves GRN reconstruction under highly sparse conditions and achieves competitive performance compared with existing approaches.
Direct prior-guided graph learning can improve the robustness and biological interpretability of GRN inference in data-limited settings and demonstrate that directed prior-guided graph learning can improve the robustness and biological interpretability of GRN inference in data-limited settings.
N. Alkhateeb, Mamoun A. Awad· Frontiers in Bioinformatics· 1 citation
Reconstruction of gene regulatory networks (GRNs) is essential for uncovering regulatory relationships between transcription factors (TFs) and target genes. With advances in single-cell RNA sequencing (scRNA-seq), cell-type-specific GRN inference has become an important direction in systems biology; however, existing deep learning methods still struggle with noisy expression, symmetric link scoring that cannot capture regulatory direction, and difficulty separating co-expression and co-regulation from direct regulation. Built upon the variational autoencoder (VAE)-graph attention network (GAT) framework of GRANet [1], we propose DMVD-GRN with enhanced VAE multi-view denoising, skew-symmetric directed decoding, and structure-aware joint decoding, integrating expression correlation, second-order adjacency, and Jaccard co-regulation priors. On 14 tasks of the STRING benchmark, DMVD-GRN achieves superior AUROC and AUPRC, with more pronounced AUPRC gains, thereby improving identification of true regulatory edges under extreme class imbalance.
PMF-GRN and GLM-Prior motivate a dual-stage view of GRN reconstruction in which sequence-derived priors provide a transferable starting scaffold and probabilistic inference refines regulatory estimates with quantified uncertainty under incomplete evaluation resources.
Understanding gene regulation at single-cell resolution is crucial for unraveling development, disease, and cellular identity. We introduce single-cell regulatory graph attention network (scReGAT), a deep learning framework that integrates prior knowledge of cis-regulatory element (cRE)-gene and transcription factor-gene interactions to reconstruct cell-specific regulatory networks. Central to scReGAT is a knowledge-guided regulatory graph (kRG), which combines experimentally validated regulatory interactions with cell-resolved chromatin accessibility profiles. These graphs serve as the foundation for training a Graph Attention Network (GAT) to predict gene expression and quantify the contribution of specific regulatory interactions using an interpretable regulatory score for each edge. In benchmarking across five single-cell multi-omics datasets, scReGAT successfully recapitulates known cell-type-specific cRE-gene interactions. In both neuroblastoma and osteogenic differentiation systems, it uncovers dynamic regulatory rewiring that predicts transcriptional transitions. Furthermore, by integrating genome-wide association studies loci from Alzheimer's disease, multiple sclerosis, and schizophrenia, scReGAT identifies disease-associated cell types and uncovers candidate regulatory mechanisms underlying complex trait associations. These results position scReGAT as a robust and generalizable framework for decoding long-range gene regulation at single-cell resolution. The source code of scReGAT can be accessed at https://github.com/TianLab-Bioinfo/scReGAT/ and https://ngdc.cncb.ac.cn/biocode/tool/BT008081.
Accurate inference of gene regulatory networks (GRNs) from single-cell gene expression data is challenging due to noise, data sparsity, and variability in gene-gene associations across cells. We propose CoReGRN (Contextual Refinement of Gene Regulatory Networks), a nonparametric, context-aware post-processing framework that refines inferred GRNs by reweighting candidate regulatory interactions using Mutual Information based association strength and local network context. The method uses empirical cumulative distribution function (ECDF) based scores to assess how unusual each interaction is relative to the connectivity patterns of the genes it connects. Then combines this contextual information with the original edge confidence scores. We evaluate the CoReGRN framework on gold-standard datasets from the BEELINE benchmark suite across multiple state-of-the-art GRN inference algorithms. The results show consistent performance improvements, with average absolute gains of 0.097 in AUROC, 0.130 in AUPR, 0.133 in MCC, and 0.095 in F1-Score. We further apply the framework to a single-cell HIV-Leishmaniasis dataset, where the refined networks support the analysis of disease-specific regulatory hubs and interactions. Comparison with existing biological knowledge identifies both known and potentially novel regulatory relationships across HIV infection, HIV-Leishmaniasis, and HIV-Visceral Leishmaniasis conditions. The analysis includes hub gene identification, interaction network analysis, and literature based validation using GeneMANIA.
E. M., Jereesh A S, G. S. Kumar· BMC Genomics· 0 citations
Gene regulatory networks (GRNs) govern cellular functions by coordinating gene expression programs. These regulatory relationships are strongly shaped by local microenvironments, giving rise to dynamic, spatially varying regulatory patterns across tissues. Therefore, it is crucial to infer GRNs at higher, cell-specific resolution while jointly modeling spatial context. However, most existing GRN inference approaches focus on cell-type–level networks or infer cell-specific GRNs without incorporating neighborhood and positional information.
We propose SVGRN, a deep learning framework for inferring spatially resolved, high-resolution GRNs from spatial transcriptomics data. SVGRN integrates gene expression, regulatory interactions, and spatial coordinates within a structural equation modeling framework implemented by a conditional variational autoencoder, to learn nonlinear, spatially varying regulatory programs in an unsupervised manner. By conditioning on target locations and incorporating neighborhood information, SVGRN refines tissue-level regulation into spot- or cell-specific GRNs. Across simulated datasets, SVGRN consistently outperforms existing methods under diverse and challenging settings. Applications to seqFISH mouse embryo data and Visium human cutaneous squamous cell carcinoma and fallopian tube datasets demonstrate that SVGRN captures spatially varying regulatory programs underlying development, tumor progression, and tissue organization, highlighting its robustness and broad applicability.
The source code and data are available at https://github.com/lyrrrr/SVGRN.
Supplementary data are available at Bioinformatics Advances online.
Yurui Li, Jin Chen, Ting Lu et al.· Bioinformatics Advances· 0 citations