Yoon Sora, Vahedi Golnaz
Department of Genetics, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA 19104, USA.
Institute for Immunology, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA 19104, USA.
bioRxiv. 2025 Jun 17:2025.06.12.659372. doi: 10.1101/2025.06.12.659372.
Three-dimensional (3D) enhancer communities are key regulators of gene expression, shaping cell fate decisions and contributing to disease pathogenesis. Assays such as H3K27ac HiChIP have been used to map enhancer-enhancer interactions and define enhancer communities; however, their reliance on antibody-based enrichment restricts scalability and cross-cell-type applicability. In contrast, Hi-C provides an unbiased, genomewide view of chromatin architecture but lacks direct annotation of regulatory elements, limiting its utility for enhancer-focused analyses. To bridge this gap, we introduce Hi-Cociety-a graph-based computational framework and accompanying R package that infers 3D enhancer communities directly from Hi-C data, without relying on histone modification or chromatin accessibility measurements. Hi-Cociety constructs a network of significant interactions and applies clustering algorithms to define chromatin interaction modules. Applying Hi-Cociety to Hi-C measurements in T lymphocytes, we identified highly connected modules enriched for active transcription, chromatin accessibility, and histone acetylation. Notably, modules identified in T cells pinpoint critical genes central to T cell biology. Hi-Cociety also detects cell-type-specific differences in chromatin organization, highlighting dynamic regulatory rewiring across T cell states. Our findings underscore the importance of network properties- connectivity, transitivity, and centrality-in shaping gene regulation through 3D genome organization. Hi-Cociety provides a scalable and versatile tool for mapping enhancer communities at scale, advancing our understanding of immune cell identity and the regulatory logic encoded in 3D chromatin structure.
三维（3D）增强子群落是基因表达的关键调节因子，塑造细胞命运决定并促成疾病发病机制。诸如H3K27ac HiChIP等检测方法已被用于绘制增强子 - 增强子相互作用图谱并定义增强子群落；然而，它们对基于抗体富集的依赖限制了可扩展性和跨细胞类型的适用性。相比之下，Hi-C提供了染色质结构的无偏全基因组视图，但缺乏调控元件的直接注释，限制了其在聚焦增强子分析中的效用。为了弥补这一差距，我们引入了Hi-Cociety——一种基于图的计算框架及配套的R包，可直接从Hi-C数据推断3D增强子群落，而无需依赖组蛋白修饰或染色质可及性测量。Hi-Cociety构建了一个显著相互作用的网络，并应用聚类算法来定义染色质相互作用模块。将Hi-Cociety应用于T淋巴细胞的Hi-C测量中，我们鉴定出高度连接的模块，这些模块富含活性转录、染色质可及性和组蛋白乙酰化。值得注意的是，在T细胞中鉴定出的模块确定了T细胞生物学核心的关键基因。Hi-Cociety还检测到染色质组织中的细胞类型特异性差异，突出了T细胞状态间动态的调控重排。我们的研究结果强调了网络属性——连通性、传递性和中心性——在通过3D基因组组织塑造基因调控中的重要性。Hi-Cociety为大规模绘制增强子群落提供了一种可扩展且通用的工具，推进了我们对免疫细胞身份以及3D染色质结构中编码的调控逻辑的理解。
