Abstract
Cell–cell communication (CCC) is central to the organization, function, and plasticity of multicellular life. Advancing experimental technologies and growing insights into complex multicellular systems and disease microenvironments are driving the demand for experimentally validated CCCs that are systematically and manually curated across diverse tissues, phenotypes, and signaling modalities. Here, we present CCCdb (http://www.licpathway.net/cccdb/index.php), a comprehensive, manually curated database of experimentally validated CCCs for human and mouse. A total of 8467 entries were extracted from thousands of publications, each annotated with standardized information on cell types, tissues, and phenotypes. These entries cover 98 tissues, 1132 cell types, and 548 phenotypes, mediated through communication via direct contact, autocrine, paracrine, and endocrine signaling. CCCdb curates experiment-supported CCCs across cellular subtypes, tissue interfaces, and physiological or pathological states. To enhance accessibility and biological interpretability, CCCdb integrates a ReAct-based AI assistant that enables intelligent natural-language queries and intuitive navigation through biological information. We believe that CCCdb will serve as a foundational resource for elucidating the mechanisms by which cells maintain tissue homeostasis and drive disease progression.
Graphical Abstract
Graphical Abstract.
Introduction
Cell–cell communication (CCC) enables biological coordination via direct contact, autocrine, paracrine, and endocrine mechanisms [1–3]. It is essential for maintaining physiological homeostasis in multicellular organisms [4]. CCCs coordinate essential physiological processes, including tissue homeostasis, development, and immunity, across spatiotemporal scales in complex multicellular systems [5, 6]. An estimated 37 trillion human and 10 trillion murine cells rely on CCCs to sustain physiological equilibrium and respond to environmental stimuli [7–9]. Dysfunctional CCCs, including aberrant interactions between cancer cells and abnormal stromal/immune cells, underlie diverse disease phenotypes [10]. Recent advancements in single-cell multi-omic technologies (scRNA-seq, scATAC-seq, spatial transcriptomics, and single-cell proteomics) have empowered researchers to resolve cellular landscapes and dissect CCCs with exceptional resolution [11, 12]. To systematically infer CCCs, several computational frameworks have been developed [13–22], the majority of predicted interactions still lack systematic biological validation [23, 24]. Experimental confirmation using low-throughput techniques, such as yeast two-hybrid assays, co-immunoprecipitation (co-IP), BioID/APEX proximity labeling, FRET imaging, and surface plasmon resonance (SPR), remains indispensable for establishing high-confidence CCCs [25, 26]. A manually curated, experimentally validated reference of CCCs across tissues, cell types, and signaling modalities is needed to support mechanistic research and computational inference tools.
Existing resources such as CITEdb [27], CCIDB [28], TICCom [29], CellTalkDB [30], MACC [31], OcmniPath [32], and so on have substantially advanced CCC research by systematically organizing experimentally reported interactions. However, the rapid expansion of experimental biology, encompassing both technological innovations and biological discoveries, presents unprecedented opportunities to advance the design of CCC databases [33]. A comprehensive, experimentally validated reference of CCC events across human and mouse is critically needed to support systematic analyses of intercellular interactions. Such a resource would enable high-confidence annotation of interacting cell types, signaling molecules, tissues, and experimental evidence across diverse phenotypic states. To ensure semantic consistency and facilitate computational reuse, interacting cell types, tissues, and physiological or pathological states should be annotated using standardized ontologies. In addition, curated CCCs should include expanded molecular context, capturing not only classical ligand–receptor pairs but also key pathway components, modulators, and cell-type–specific marker genes. Collecting experimentally validated CCCs provides a critical foundation for interpreting diverse high-throughput technologies and advancing algorithm development, while their systematic integration into CCC resources will be essential for building comprehensive, high-resolution models of intercellular communication. Finally, to enhance usability, modern CCC resources should support advanced semantic search and AI-assisted query systems, enabling efficient biological interpretation.
Here, we present CCCdb (http://www.licpathway.net/cccdb/index.php), a comprehensive, manually curated database of experimentally validated CCCs in human and mouse. CCCdb is curated by domain experts to systematically compile high-confidence CCC events, elucidate functional association (e.g. marker genes, ligand–receptor pairs, signaling pathways), and document supporting low-throughput experimental validation. It is built upon a meticulous manual curation of literature published between 1973 and 2025. The current release includes 8467 curated entries, each annotated with 25 attributes such as cell types, tissues, phenotypes, and experimental methods. After consolidating redundant annotations, this corresponds to 3000 unique CCC interactions across 98 tissues, 1132 cell types, and 548 phenotypic states. These communications operate through four primary modes of intercellular signaling (direct contact [34], autocrine [35], paracrine [36], and endocrine [37]). To standardize entity representation and facilitate interoperability, CCCdb uses reference ontologies: each cell type is assigned a Cell Ontology (CL) identifier (e.g. CL_0000235 for Macrophage), each tissue is annotated using the Uberon ontology (e.g. UBERON_0001637 for Artery), and each disease association is mapped to Disease Ontology (e.g. DOID:769 for Neuroblastoma). CCCdb also features a ReAct-based AI query engine, allowing users to pose biological questions in natural language, which are automatically translated into SQL for structured query execution [38]. We anticipate that CCCdb will serve as a foundational resource, advancing multi-omics algorithm development and enabling deeper mechanistic insights into CCCs in both health and disease.
Materials and methods
Manual curation of experimentally validated cell communication
Step 1: Comprehensive retrieval of CCC-related publications using targeted search strategies
To systematically compile experimentally validated CCC events, a multistep literature retrieval strategy was implemented. Our approach was inspired by the methodologies of CellMarker 2.0 [39], TF-Marker [40], and ENdb [41]. We designed three complementary keyword-based strategies to ensure exhaustive collection of CCC-related studies: Strategy A: A broad PubMed (https://pubmed.ncbi.nlm.nih.gov/) search was performed using canonical CCC-related terms, including “cell communication”, “cell interaction”, “cell contact”, “cell crosstalk”, “cell adhesion”, “extracellular vesicle”, and “gap junction communication”, retrieving ∼72 000 publications. Strategy B: Over 25 Boolean combinations of CCC-related concepts, such as “cell communication AND ligand–receptor” or “cell interaction AND autocrine”, were applied to capture more complex phrasing, yielding roughly 46 000 publications. Strategy C: To include CCC events reported in alternative biological contexts, additional keywords, including “cell migration”, “cell invasion”, and “connexon” were employed, resulting in ∼60 000 publications.
Collectively, after deduplication, around 100 000 publications were retrieved covering the earliest PubMed records to 2025.
Step 2: Refinement by publication type, title/abstract filtering, and species restriction
To focus on biologically validated CCC events, publications retrieved in Step 1 were classified by type to distinguish primary research publications reporting novel experimental CCCs from review publications that summarized experimentally validated CCC events. Next, publications were filtered to include only those containing CCC-related keywords in both the title and abstract, ensuring topic relevance. Finally, species restriction was applied to retain publications conducted in humans or mouse, the primary models for mechanistic and translational CCC research. After applying these criteria, the publication corpus contained ∼30 000 candidate publications. Each publication was annotated with metadata including its type, species, and the presence of CCC-related keywords in the title and abstract, providing a structured foundation for subsequent manual curation.
Step 3: Manual evaluation and expert-level extraction of CCC evidence
Full texts of publications identified in Step 2 were subjected to comprehensive manual curation. A dedicated team of trained curators systematically evaluated a large set of publications, extracting experimentally validated CCC events along with 25 associated attributes, including interacting cell types, important signaling or protein pairs, tissues and phenotypes, experimental validation methods, PMID, and title. CCCs supported by direct biological evidence—such as co-IP, RNAscope, BioID, FRET, or spatial transcriptomics—were cataloged. Curators carefully examined CCC information embedded in figures, tables, and supplementary materials to ensure complete capture of mechanistic details. Each publication was independently reviewed by multiple curators to maintain accuracy, reproducibility, and consistency of annotation. This rigorous pipeline ensured that the resulting CCC database reflects high-confidence, biologically validated interactions across tissues and cell types (Fig. 1I and Supplementary Fig. S1).
Figure 1.
The overall design of CCCdb. (I-II) The data sources and the workflow of the database, displaying fundamental statistical information about the data. (III) The main pages of CCCdb, including ‘Home’, ‘Search’, ‘Browse’, ‘Analysis’, ‘Statistics’, ‘Submit’ and ‘Download’.
Processing and annotation
We manually curated essential biological attributes for each experimentally validated CCC entry, including species (human or mouse), tissue type, cell type, key signaling protein(s), phenotype name, experiment type (low-throughput, low and high-throughput), experiment name, pathways and detailed publication information (e.g. title, PubMed ID, and publication year). To avoid redundancy, entries with the same CCC pair, cell type, tissues, species, and phenotype were manually reviewed and grouped. CCC events were classified into four major modes of intercellular communication (direct contact [34], autocrine [35], paracrine [36], and endocrine [37]), and the supporting literature for each interaction was systematically documented.
To harmonize nomenclature across studies and ensure semantic consistency, we standardized tissue, cell type, and disease phenotype terms using established biomedical ontologies. Specifically, tissue names were aligned to the Uberon anatomical ontology [42], cell types were normalized according to the Cell Ontology [43], and disease-related phenotypes were mapped to the Disease Ontology [44]. To support precise entity representation and enhance interoperability across biological databases, each CCC entry was annotated with standardized ontology identifiers: CL IDs for cell types (e.g. CL:0000235 for macrophages), UBERON IDs for tissues (e.g. UBERON:0001637 for arteries), and DOIDs for disease phenotypes (e.g. DOID:769 for neuroblastoma) (Fig. 1II).
All curated CCC entries and their associated annotations were programmatically integrated into a structured MySQL relational database. An overview of the database architecture, data integration framework, and key features is depicted in Fig. 1.
Database implementation
The current implementation of CCCdb is built on MySQL (v5.7.27) and hosted on a Linux-based Apache web server. Server-side functionality is supported by PHP (v5.6.40), while the user interface is constructed using Bootstrap (v3.3.7) and jQuery (v2.1.1). Interactive data visualizations are rendered using ECharts. For optimal display and compatibility, modern browsers that fully support HTML5—such as Firefox, Google Chrome, or Safari—are recommended. CCCdb is freely accessible to the scientific community without the need for user registration or login. The online portal is available at http://www.licpathway.net/cccdb/. Source code for the PHP framework is publicly available on GitHub (https://github.com/TOSTRING-Z/cccdb), and the complete dataset can be accessed via https://zenodo.org/records/16877264.
Description of the database
Overview of CCCdb
CCCdb provides experimentally validated CCCs manually curated from 1656 publications, supported by low-throughput assays including co-IP, BioID, APEX, FRET imaging, SPR, and enzyme-linked immunosorbent assay (ELISA). Each entry includes standardized annotations such as cell names, tissue types, phenotypes, communication modes, experimental types and names, pathway names, PMIDs, paper types, and publication years. All biological terms are standardized using official ontologies Uberon_ID [42], CL_ID [43], DOID [44], and KEGG_ID [45] to ensure consistency across datasets from diverse sources. The data are uniformly annotated to support comparative analysis of CCC types, signaling pathways, diseases, and experimental evidence. CCCdb comprises 3000 CCC interactions involving 1132 cell types, 98 tissues, 48 pathways, and 548 phenotypes, categorized into four communication modes: direct contact, autocrine, paracrine, and endocrine.
A key feature of CCCdb is the integration of an AI agent, which enables the database is organized into four modules, offering four search methods, an interactive browser interface for data exploration and download, a statistics dashboard for global data overview, and a user submission portal for contributing new findings and feedback (Fig. 1III).
Browse interface for CCCdb
The ‘Browse’ page of CCCdb features an interactive, paginated table that allows users to efficiently explore experimentally validated CCCs. Users can apply customized filters based on ‘Species’, ‘Communication Type’, ‘Experiment Type’, ‘Tissue Type’, and ‘Phenotype’. The table displays detailed information for each interaction, including the interacting cell types, tissue/sub-tissue types, species, phenotype, and the corresponding PubMed ID. The number of entries shown per page can be modified using the ‘Entries per page’ drop-down menu. In addition, each record is hyperlinked to ‘CCC Details’, enabling direct access to supporting experimental evidence (Fig. 2A).
Figure 2.
A schematic workflow of CCCdb. (A) The web images in the home page allow to quick search for cell communications in different tissues. (B) The ‘Browse’, ‘Search’, and ‘Analysis’ allow the users to browse and search cell communications. (C) AI assistant tool for understanding cell communications. (D) The details information for the users interested CCC.
Search interface for CCCdb
CCCdb provides four complementary search modes, including “Search by Cell to Cell Type”, “Search by communication modes”, “Search by Tissue Type”, and “Search by Disease Type”. In the Disease Type-based search, users can specify the species and disease types of interest to filter the list of CCCs. The result table displays key information, including cell name, species, tissue, phenotype, and PMID. Users can click “more details” to obtain further detailed information (Fig. 2B).
AI assistant analysis for CCCdb
CCCdb integrates an AI assistant that enables users to query CCCs using natural language, such as “paracrine signaling in Giloma” or “communications between T cell and Malignant cell”. Leveraging advanced natural language processing techniques, the assistant returns structured results accompanied by relevant information. Users can customize the number of results returned and access example queries to facilitate rapid exploration.
Moreover, users can simply enter keywords of interest to analyze. The AI assistant then returns structured results with supporting evidence. For example, users enter “glioma” as keyword; then by clicking “More details”, users can further view experimental references, molecular mechanisms, and specific communications (Fig. 2C). Results showed that malignant cells communicate with T cells via paracrine signaling. Low-throughput experiments, including quantitative real-time polymerase chain reaction (qRT-PCR), immunohistochemistry (IHC), and flow cytometry, support the involvement of the chemokine CXCL16 and its receptor CXCR6 (Fig. 2D). These results have been reported (PMID: 27784296) [46], and the experimental system corresponds to human brain tissue. This example illustrates how glioma cells modulate immune responses through chemokine-mediated paracrine interactions, identifying potential therapeutic targets. Overall, this AI-driven feature enhances data accessibility and interpretability, allowing researchers to efficiently uncover complex intercellular communications across diverse biological processes [47, 48].
In addition, CCCdb records 48 signaling pathways that mediate CCCs. Through the AI-assisted interface, users can directly query by entering a specific pathway name. For example, by inputting MAPK into the search box and clicking Analyze, the AI assistant will return all CCCs associated with the MAPK signaling pathway.
Results
Database statics
The current version of CCCdb includes 5276 paracrine, 2811 direct contact, 226 autocrine, and 154 endocrine interactions (Fig. 3A). The majority of cells communicate predominantly through paracrine signaling. The top 10 phenotypes ranked by entry count include the normal phenotype, breast carcinoma, and melanoma, among others (Fig. 3B). In normal phenotype, direct contact signaling occurs most frequently between B cells and T cells (Fig. 3C). In breast carcinoma, cancer cells most frequently communicate with endothelial and epithelial cells (Fig. 3D). CCCdb contains 1656 publications, 3000, cell pairs, and 548 phenotypes (Fig. 3E).
Figure 3.
Statistics of cell communications in CCCdb. (A) Distribution of four functional communication modes (paracrine, direct contact, and endocrine). (B) Top phenotypes ranked by number of curated entries. (C) Most frequent cell communications in normal phenotypes ranked by number of curated entries. (D) Cell communications associated with breast carcinoma and their corresponding communication modes. (E) Comparison of CCCdb with existing resources in curated publications, cell pairs, phenotypes, and communication modes.
Case study: Functionally relevant immune–tumor interactions in breast carcinoma.
Breast carcinoma represents not only a malignancy of epithelial origin but also a dynamic multicellular ecosystem shaped by persistent immune–tumor interactions [49]. Despite advances in immunotherapy—such as checkpoint blockade and antibody-drug conjugates—therapeutic resistance and heterogeneous clinical responses remain major challenges [50, 51]. Emerging studies suggest that the functional crosstalk between immune cells and tumor cells, rather than immune presence alone, critically determines treatment outcomes [52]. However, these interactions often occur within spatially confined niches and rely on complex signaling modalities that are difficult to infer solely from high-throughput transcriptomic data.
To elucidate functionally relevant CCCs in breast carcinoma, we queried CCCdb with a focus on experimentally validated interactions within the tumor immune microenvironment (Fig. 4A). A notable example, supported by multimodal experimental validation ‘PMID: 40055573’ [53], revealed that neutrophils establish perivascular contact clusters with tumor cells, forming a spatially organized signaling niche. In this context, neutrophils engage in direct physical interactions with epithelial tumor cells, promoting a signaling microenvironment that enhances tumor aggressiveness [54] (Fig. 4B). Mechanistically, this interaction is partially mediated by the VEGFA–VEGFR1, a well-established driver of angiogenesis and tumor progression. Beyond breast carcinoma, VEGFA and VEGFR1 are extensively implicated in vascular remodeling, immune cell recruitment, and modulation of the tumor microenvironment across diverse cancer types (Fig. 4C). These neutrophil–tumor cell communications illustrate how innate immune cells can be co-opted to support malignancy, highlighting potential therapeutic strategies (Fig. 4D). Disruption of VEGFA–VEGFR1 signaling or interference with neutrophil–tumor cell contacts may potentiate the efficacy of immunotherapies and anti-angiogenic treatments [55].
Figure 4.
Functionally relevant immune–tumor interactions in breast carcinoma. (A) AI assistant tool for understanding cell communications in breast carcinoma. (B) The details information for the neutrophil and epithelial cell communication in breast carcinoma. (C) The important proteins in CCCs. (D) The CCCs of breast carcinoma in CCCdb.
Discussion
Cell communication is fundamental to multicellular organization, driving key biological processes from embryonic development to tumorigenesis [56]. Advances in single-cell and spatial omics have enabled systematic investigation of CCCs across diverse tissue types [57, 58]. To elucidate the coordination and hierarchy among distinct cell populations during organ development and tissue remodeling, more and more computational frameworks have been developed to infer CCCs from high-throughput omics data [59]. For instance, CellChat uses social network theory and non-negative matrix factorization to identify communication patterns and key signaling pathways, while CellPhoneDB applies mean-based, permutation, and differential gene-driven methods for interaction inference. Their development and evaluation rely on experimentally validated, gold-standard interactions. At the molecular level, assays such as Y2H, co-IP, BioID, APEX, FRET, SPR, and ELISA probe distinct aspects of intercellular signaling and collectively support mechanistic analysis of CCC across biological systems [6]. Accordingly, we developed CCCdb: an expert-curated resource systematically cataloging high-confidence CCC events, their associated molecular mechanisms (e.g. marker genes, ligand–receptor interactions, signaling pathway activation), and supporting evidence from low-throughput biological assays. CCCdb provides high-confidence, experimentally supported data on cell–cell interactions, thereby serving as a reliable resource to support these computational tools.
CCCdb is an expert-curated, interactive database featuring an AI agent that facilitates the interpretation of CCC based on experimentally validated interactions. By comparing with existing databases, CCCdb has the following advantages (Table 1): (i) Greater CCC information: CCCdb provides 8467 CCC entries, which is ∼16-fold higher than CCIDB (515 entries) and 11-fold higher than TICCom (738 entries), offering a substantially richer resource for intercellular interaction analysis. (ii) Standardized biomedical annotations: CCCdb uniquely provides structured annotations using official biomedical ontologies, including CL_ID, DOID and UBERON_ID. (iii) Phenotype-resolved experimental validation: CCCdb annotates 577 phenotypes integrates evidence from low-throughput (e.g. co-IP, BioID, APEX, FRET imaging, SPR, and ELISA). (iv) AI assistant query system and modular access: CCCdb incorporates an AI agent-driven query engine that allows users to interactively access relevant CCC records via semantic search and SQL generation, enhancing data usability for both biologists and computational researchers. (v) Beyond L-R interactions, CCCdb systematically curates experimentally supported marker genes and signaling pathways involved in CCCs. (vi) Additionally, CCCdb classifies CCC into four modes—direct contact, autocrine, paracrine, and endocrine, helping researchers to understand the tissue-specific of CCC.
Table 1.
Advantages of CCCdb compared with existing CCC databases
| CCCdb | CITEdb | CCIDB | TICCOM | |
|---|---|---|---|---|
| The number of publications | 1656 | 574 | 272 | 337 |
| Total number of entries | 8467 | Not available | 515 | 738 |
| Year | 1973–2025 | 2013–2022 | 2016–2021 | Not available–2019 |
| The number of cell pairs | 3000 | 728 | 300 | 252 |
| The number of cell types | 1132 | Not available | 171 | 16 |
| CL_ID | Yes | Not available | No | No |
| The number of Tissues | 98 | Not available | 46 | No |
| UBERON_ ID | Yes | Not available | No | No |
| Phenotype | Yes | Yes | Yes | Yes |
| Disease_id | Yes | No | No | No |
| The number of phenotypes | 548 | 204 | 186 | 71 |
| AI agent analysis | Yes | No | No | No |
| Low-throughput biological assays | Yes | Yes | Yes | Yes |
| Experiment name | Yes | Not available | No | Yes |
We are committed to the continuous maintenance and upgrade of CCCdb, ensuring its long-term relevance and utility for the scientific community. (i) We will continually maintain and enrich CCCdb by adding newly confirmed CCC events across emerging cell subtypes, disease settings, and various experimental models such as 3D organoid cultures and spatial transcriptomic platforms, ensuring alignment with advancing experimental techniques and biological discoveries. (ii) We encourage community engagement by providing an online submission portal, allowing users to contribute CCC events of interest. These submissions will inform future curation and help tailor the database to evolving research needs. (iii) A user feedback system has also been implemented to collect suggestions, report issues, and gather perspectives. By actively considering users’ advice, we strive to enhance the platform’s usability, accessibility, and scientific value. (iv) CCCdb integrates an AI-assisted query engine, enabling users to interactively retrieve curated CCC entries using natural language questions. In the future, we plan to advance this AI module to achieve automated surveillance of newly published CCC-related literature on PubMed, thereby enabling CCCdb to dynamically assimilate the latest experimentally validated findings. This next-generation, near real-time, AI-driven curation framework will empower both experimental biologists and computational researchers, accelerating hypothesis generation and fostering timely scientific discovery.
In summary, we are dedicated to ensuring that CCCdb remains a dynamic, high-quality, and user-friendly resource that supports the research community in uncovering intercellular communication and addressing a broad range of biological questions.
Supplementary Material
Acknowledgements
Author contributions: Mingcong Xu (Data curation [equal], Investigation [equal], Methodology [equal], Writing - original draft [equal]), Guorui Zhang (Data curation [equal], Methodology [equal]), Xuan Wang (Data curation [equal], Investigation [equal]), Yiqing Chen (Visualization [equal]), Chenchen Feng (Methodology [equal]), Jincheng Guo (Formal analysis [equal]), Xuan Fan (Formal analysis [equal]), Liyuan Liu (Investigation [equal]), Yuezhu Wang (Resources [equal]), Ting Cui (Visualization [equal]), Jiaqi Liu (Formal analysis [equal]), Libo Luo (Investigation [equal]), Qing Xun (Visualization [equal]), Yiguang Fan (Data curation [equal]), Xiaoyu Ma (Visualization [equal]), Huifang Tang (Project administration [equal], Writing—original draft [equal], Writing—review & editing [equal]), Chunquan Li (Project administration [equal], Visualization [equal], Writing—original draft [equal], Writing—review & editing [equal]), and Desi Shang (Data curation [equal], Project administration [lead], Writing—original draft [equal], Writing—review & editing [equal])
Contributor Information
Mingcong Xu, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China; School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080, China.
Guorui Zhang, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Xuan Wang, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Yiqing Chen, School of Traditional Chinese Medicine, Beijing University of Chinese Medicine, Beijing 1000209, China.
Chenchen Feng, School of Computer, University of South China, Hengyang, Hunan 421001, China.
Jincheng Guo, School of Traditional Chinese Medicine, Beijing University of Chinese Medicine, Beijing 1000209, China.
Xuan Fan, School of Traditional Chinese Medicine, Beijing University of Chinese Medicine, Beijing 1000209, China.
Liyuan Liu, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China; Department of Biochemistry and Molecular Biology, School of Basic Medical Sciences, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Yuezhu Wang, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Ting Cui, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Jiaqi Liu, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Libo Luo, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Qing Xun, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Yiguang Fan, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Xiaoyu Ma, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Huifang Tang, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China; School of Computer, University of South China, Hengyang, Hunan 421001, China.
Chunquan Li, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China; School of Computer, University of South China, Hengyang, Hunan 421001, China; Department of Biochemistry and Molecular Biology, School of Basic Medical Sciences, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China; Hunan Provincial Maternal and Child Health Care Hospital, National Health Commission Key Laboratory of Birth Defect Research and Prevention, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China; Key Laboratory of Rare Pediatric Diseases, Ministry of Education, University of South China, Hengyang, Hunan 421001, China.
Desi Shang, The First Affiliated Hospital & Hunan Provincial Key Laboratory of Multi-omics and Artificial Intelligence of Cardiovascular Diseases, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China; School of Computer, University of South China, Hengyang, Hunan 421001, China; Department of Biochemistry and Molecular Biology, School of Basic Medical Sciences, Hengyang Medical School, University of South China, Hengyang, Hunan 421001, China.
Supplementary data
Supplementary data is available at NAR online.
Conflict of interest
None declared.
Funding
National Natural Science Foundation of China (62171166, 62272211, 62572223, 82570463); The Science and Technology Innovation Program of Hunan Province (2024RC1062); Noncommunicable Chronic Diseases-National Science and Technology Major Project (2024ZD0528200); The Innovation Platform and Talent Program (2023TP1047); Natural Science Foundation of Hunan Province (2023JJ30536, 2023JJ30535); Hunan Provincial Health High-Level Talent Scientific Research Project (R2023131); Health Research Project of Hunan Provincial Health Commission (W20241008); Clinical Research 4310 Program of the University of South China (20224310NHYCG05, 20214310NHYCG03).
Data availability
CCCdb is freely available online at http://www.licpathway.net/cccdb/index.php, and there is no login requirement.
References
- 1. Domínguez Conde C, Xu C, Jarvis LBet al. Cross-tissue immune cell analysis reveals tissue-specific features in humans. Science. 2022;376:eabl5197. 10.1126/science.abl5197. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Shi Q, Chen Y, Li Yet al. Cross-tissue multicellular coordination and its rewiring in cancer. Nature. 2025;643:529–38. 10.1038/s41586-025-09053-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Luca BA, Steen CB, Matusiak Met al. Atlas of clinically distinct cell states and ecosystems across human solid tumors. Cell. 2021;184:5482–5496.e28. 10.1016/j.cell.2021.09.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Zhang P, Li S. Human cross-tissue cell atlases: unprecedented resources towards systematic understanding of physiology and diseases. Signal Transduct Target Ther. 2022;7:352. 10.1038/s41392-022-01201-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Tang F, Li J, Qi Let al. A pan-cancer single-cell panorama of human natural killer cells. Cell. 2023;186:4235–4251.e20. 10.1016/j.cell.2023.07.034. [DOI] [PubMed] [Google Scholar]
- 6. Su J, Song Y, Zhu Zet al. Cell–cell communication: new insights and clinical implications. Signal Transduct Target Ther. 2024;9:196. 10.1038/s41392-024-01888-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Sender R, Fuchs S, Milo R. Revised estimates for the number of Human and bacteria cells in the body. PLoS Biol. 2016;14:e1002533. 10.1371/journal.pbio.1002533. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Milo R, Jorgensen P, Moran Uet al. BioNumbers – the database of key numbers in molecular and cell biology. Nucleic Acids Res. 2010;38:D750–753. 10.1093/nar/gkp889. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Bidanta S, Börner K, Herr Ii BWet al. Functional tissue units in the Human Reference Atlas. Nat Commun. 2025;16:1526. 10.1038/s41467-024-54591-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Kurche JS, Stancil IT, Michalski JEet al. Dysregulated cell–cell communication characterizes pulmonary fibrosis. Cells. 2022;11:3319. 10.3390/cells11203319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Agrawal A. Decoding cell–cell communication using spatial transcriptomics. Nat Rev Genet. 2025;26:295. 10.1038/s41576-025-00824-3. [DOI] [PubMed] [Google Scholar]
- 12. Vandereyken K, Sifrim A, Thienpont Bet al. Methods and applications for single-cell and spatial multi-omics. Nat Rev Genet. 2023;24:494–515. 10.1038/s41576-023-00580-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Yang W, Wang P, Xu Set al. Deciphering cell–cell communication at single-cell resolution for spatial transcriptomics with subgraph-based graph attention network. Nat Commun. 2024;15:7101. 10.1038/s41467-024-51329-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Ding Q, Yang W, Xue Get al. Dimension reduction, cell clustering, and cell–cell communication inference for single-cell transcriptomics with DcjComm. Genome Biol. 2024;25:241. 10.1186/s13059-024-03385-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Yang W, Wang P, Luo Met al. DeepCCI: a deep learning framework for identifying cell–cell interactions from single-cell RNA sequencing data. Bioinformatics. 2023;39:btad596. 10.1093/bioinformatics/btad596. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Zhang Y, Liu T, Wang Jet al. Cellinker: a platform of ligand-receptor interactions for intercellular communication analysis. Bioinformatics. 2021;37:2025–32. 10.1093/bioinformatics/btab036. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Efremova M, Vento-Tormo M, Teichmann SAet al. CellPhoneDB: inferring cell–cell communication from combined expression of multi-subunit ligand–receptor complexes. Nat Protoc. 2020;15:1484–506. 10.1038/s41596-020-0292-x. [DOI] [PubMed] [Google Scholar]
- 18. Jin S, Guerrero-Juarez CF, Zhang Let al. Inference and analysis of cell–cell communication using CellChat. Nat Commun. 2021;12:1088. 10.1038/s41467-021-21246-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Moosavi SH, Eide PW, Eilertsen IAet al. De novo transcriptomic subtyping of colorectal cancer liver metastases in the context of tumor heterogeneity. Genome Med. 2021;13:143. 10.1186/s13073-021-00956-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Ma Q, Li Q, Zheng Xet al. CellCommuNet: an atlas of cell–cell communication networks from single-cell RNA sequencing of human and mouse tissues in normal and disease states. Nucleic Acids Res. 2024;52:D597–606. 10.1093/nar/gkad906. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Zhang Y, Liu T, Hu Xet al. CellCall: integrating paired ligand–receptor and transcription factor activities for cell–cell communication. Nucleic Acids Res. 2021;49:8520–34. 10.1093/nar/gkab638. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Dupuis L, Debeaupuis O, Simon Fet al. CausalCCC: a web server to explore intracellular causal pathways enabling cell–cell communication. Nucleic Acids Res. 2025;53:W125–31. 10.1093/nar/gkaf404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Armingol E, Baghdassarian HM, Lewis NE. The diversification of methods for studying cell–cell interactions and communication. Nat Rev Genet. 2024;25:381–400. 10.1038/s41576-023-00685-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Fan X, He Z, Guo Jet al. Leveraging TME features and multi-omics data with an advanced deep learning framework for improved cancer survival prediction. Sci Rep. 2025;15:14282. 10.1038/s41598-025-98565-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Mohammed H, Carroll JS. Approaches for assessing and discovering protein interactions in cancer. Mol Cancer Res. 2013;11:1295–302. 10.1158/1541-7786.MCR-13-0454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Bosch JA, Chen C-L, Perrimon N. Proximity-dependent labeling methods for proteomic profiling in living cells: an update. Wiley Interdiscip Rev Dev Biol. 2021;10:e392. 10.1002/wdev.392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Shan N, Lu Y, Guo Het al. CITEdb: a manually curated database of cell–cell interactions in human. Bioinformatics. 2022;38:5144–8. 10.1093/bioinformatics/btac654. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Noh JY, Lee HI, Choi J-Het al. CCIDB: a manually curated cell–cell interaction database with cell context information. Database (Oxford). 2023;2023:baad057. 10.1093/database/baad057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Xie Y, Zhou W, Shi Jet al. A global database for modeling tumor-immune cell communication. Sci Data. 2023;10:444. 10.1038/s41597-023-02342-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Shao X, Liao J, Li Cet al. CellTalkDB: a manually curated database of ligand–receptor interactions in humans and mice. Brief Bioinform. 2021;22:bbaa269. 10.1093/bib/bbaa269. [DOI] [PubMed] [Google Scholar]
- 31. Gao J, Mo S, Wang Jet al. MACC: a visual interactive knowledgebase of metabolite-associated cell communications. Nucleic Acids Res. 2024;52:D633–9. 10.1093/nar/gkad914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Türei D, Korcsmáros T, Saez-Rodriguez J. OmniPath: guidelines and gateway for literature-curated signaling pathway resources. Nat Methods. 2016;13:966–7. [DOI] [PubMed] [Google Scholar]
- 33. Zhang S, Zhao H, Liu Zet al. Monitoring of cell–cell communication and contact history in mammals. Science. 2022;378:eabo5503. 10.1126/science.abo5503. [DOI] [PubMed] [Google Scholar]
- 34. Gong B, Johnston JD, Thiemicke Aet al. Endoplasmic reticulum-plasma membrane contact gradients direct cell migration. Nature. 2024;631:415–23. 10.1038/s41586-024-07527-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Lee S, Chen TT, Barber CLet al. Autocrine VEGF signaling is required for vascular homeostasis. Cell. 2007;130:691–703. 10.1016/j.cell.2007.06.054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Acharyya S, Oskarsson T, Vanharanta Set al. A CXCL1 paracrine network links cancer chemoresistance and metastasis. Cell. 2012;150:165–78. 10.1016/j.cell.2012.04.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Šestan M, Raposo B, Rendas Met al. Neuronal-ILC2 interactions regulate pancreatic glucagon and glucose homeostasis. Science. 2025;387:eadi3624. [DOI] [PubMed] [Google Scholar]
- 38. Mao Y, Shangguan D, Huang Qet al. Emerging artificial intelligence-driven precision therapies in tumor drug resistance: recent advances, opportunities, and challenges. Mol Cancer. 2025;24:123. 10.1186/s12943-025-02321-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Hu C, Li T, Xu Yet al. CellMarker 2.0: an updated database of manually curated cell markers in human/mouse and web tools based on scRNA-seq data. Nucleic Acids Res. 2023;51:D870–6. 10.1093/nar/gkac947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Xu M, Bai X, Ai Bet al. TF-Marker: a comprehensive manually curated database for transcription factors and related markers in specific cell and tissue types in human. Nucleic Acids Res. 2022;50:D402–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Bai X, Shi S, Ai Bet al. ENdb: a manually curated database of experimentally supported enhancers for human and mouse. Nucleic Acids Res. 2020;48:D51–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. Mungall CJ, Torniai C, Gkoutos GVet al. Uberon, an integrative multi-species anatomy ontology. Genome Biol. 2012;13:R5. 10.1186/gb-2012-13-1-r5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43. Diehl AD, Meehan TF, Bradford YMet al. The Cell Ontology 2016: enhanced content, modularization, and ontology interoperability. J Biomed Semantics. 2016;7:44. 10.1186/s13326-016-0088-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Schriml LM, Mitraka E, Munro Jet al. Human Disease Ontology 2018 update: classification, content and workflow expansion. Nucleic Acids Res. 2019;47:D955–62. 10.1093/nar/gky1032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Kanehisa M, Furumichi M, Sato Yet al. KEGG: biological systems database as a model of the real world. Nucleic Acids Res. 2025;53:D672–7. 10.1093/nar/gkae909. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Hattermann K, Bartsch K, Gebhardt HHet al. ‘Inverse signaling’ of the transmembrane chemokine CXCL16 contributes to proliferative and anti-apoptotic effects in cultured human meningioma cells. Cell Commun Signal. 2016;14:26. 10.1186/s12964-016-0149-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Yang Z, Zeng X, Zhao Yet al. AlphaFold2 and its applications in the fields of biology and medicine. Signal Transduct Target Ther. 2023;8:115. 10.1038/s41392-023-01381-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Gao S, Fang A, Huang Yet al. Empowering biomedical discovery with AI agents. Cell. 2024;187:6125–51. 10.1016/j.cell.2024.09.022. [DOI] [PubMed] [Google Scholar]
- 49. Zhang Y, Chen H, Mo Het al. Distinct cellular mechanisms underlie chemotherapies and PD-L1 blockade combinations in triple-negative breast cancer. Cancer Cell. 2025;43:446–463.e7. 10.1016/j.ccell.2025.01.007. [DOI] [PubMed] [Google Scholar]
- 50. Zhang Y, Zhang Z. The history and advances in cancer immunotherapy: understanding the characteristics of tumor-infiltrating immune cells and their therapeutic implications. Cell Mol Immunol. 2020;17:807–21. 10.1038/s41423-020-0488-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Baxevanis CN, Fortis SP, Perez SA. The balance between breast cancer and the immune system: challenges for prognosis and clinical benefit from immunotherapies. Semin Cancer Biol. 2021;72:76–89. 10.1016/j.semcancer.2019.12.018. [DOI] [PubMed] [Google Scholar]
- 52. Lopez de Rodas M, Villalba-Esparza M, Sanmamed MFet al. Biological and clinical significance of tumour-infiltrating lymphocytes in the era of immunotherapy: a multidimensional approach. Nat Rev Clin Oncol. 2025;22:163–81. 10.1038/s41571-024-00984-x. [DOI] [PubMed] [Google Scholar]
- 53. Camargo S, Moskowitz O, Giladi Aet al. Neutrophils physically interact with tumor cells to form a signaling niche promoting breast cancer aggressiveness. Nat Cancer. 2025;6:540–58. 10.1038/s43018-025-00924-3. [DOI] [PubMed] [Google Scholar]
- 54. Xiong S, Dong L, Cheng L. Neutrophils in cancer carcinogenesis and metastasis. J Hematol Oncol. 2021;14:173. 10.1186/s13045-021-01187-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55. Zhang Q, Lu S, Li Tet al. ACE2 inhibits breast cancer angiogenesis via suppressing the VEGFa/VEGFR2/ERK pathway. J Exp Clin Cancer Res. 2019;38:173. 10.1186/s13046-019-1156-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56. Dixson AC, Dawson TR, Di Vizio Det al. Context-specific regulation of extracellular vesicle biogenesis and cargo selection. Nat Rev Mol Cell Biol. 2023;24:454–76. 10.1038/s41580-023-00576-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57. Qi X, Zhao L, Tian Cet al. Predicting transcriptional responses to novel chemical perturbations using deep generative model for drug discovery. Nat Commun. 2024;15:9256. 10.1038/s41467-024-53457-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58. Baysoy A, Bai Z, Satija Ret al. The technological landscape and applications of single-cell multi-omics. Nat Rev Mol Cell Biol. 2023;24:695–713. 10.1038/s41580-023-00615-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59. Hu Y, Rong J, Xu Yet al. Unsupervised and supervised discovery of tissue cellular neighborhoods from cell phenotypes. Nat Methods. 2024;21:267–78. 10.1038/s41592-023-02124-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
CCCdb is freely available online at http://www.licpathway.net/cccdb/index.php, and there is no login requirement.