Interactive Multi-Omic Platform for Exploring the Transcriptomic Architecture, Trajectory Dynamics, and Spatial Tissue Compartmentalization of Neutrophil Hubs.
Lead Architect & Developer: Maxence Tricaud (mtricaud.cetri@gmail.com)
NeuMapp Explorer (maintained in this repository under Transcriptomic-Explorer) is an interactive multi-omic computational platform engineered for investigating the transcriptomic landscape, developmental trajectory dynamics, and tissue compartmentalization architecture of neutrophil functional states. The suite bridges single-cell transcriptomic resolution (scRNA-seq) with tissue-level anatomical hubs, combining an interactive R/Shiny visualization frontend with an optimized scientific Python computing backend for trajectory inference and spatial niche deconvolution.
┌─────────────────────────────────────────────────────────────┐
│ NeuMapp Explorer Architecture │
└──────────────────────────────┬──────────────────────────────┘
│
┌───────────────────────────────┴───────────────────────────────┐
▼ ▼
┌─────────────────────────┐ ┌─────────────────────────┐
│ R/Shiny User Frontend │ │ Python Analytical Core │
├─────────────────────────┤ ├─────────────────────────┤
│ • modules/neumapp_ │ │ • spatial_ │
│ spatial/ (3400+ loc) │ │ deconvolution.py │
│ • modules/seurat_ │◄────────── Cross-Modality ──────────►│ • trajectory_ │
│ singlecell/ │ Coordination │ inference.py │
│ • launch_suite.R │ │ • pipeline_ │
│ • root app.R (Seurat v5)│ │ integrative.py │
└─────────────────────────┘ └─────────────────────────┘
In neutrophil immunobiology, "spatial" architecture does not merely denote 2D pixel coordinates on an artificial glass slide, but fundamentally refers to the anatomical compartmentalization and organ-specific vascular hubs pioneered by Andrés Hidalgo et al. (Cell 2019, Cell 2020, Nature Immunology 2021):
- Bone Marrow Precursor & Retention Niche: Pro-Neu and pre-Neu proliferative reservoirs (
c-Kit+,CXCR4+). - Circulating Vascular Pool: Diurnally oscillating, mature effector neutrophils (
CD62L_high,CXCR2+). - Marginal Vascular Pools: Non-canonical intravascular marginated reservoirs adhering in lung capillary beds and splenic red pulp.
- Target Tissue & Microenvironment Hubs: Extravasated neutrophil phenotypes adapting to localized niches, including acute inflammatory foci and tumor microenvironments (TANs: N1 anti-tumor vs N2 immunosuppressive niches).
- Continuous Lineage Kinetics (scRNA-seq): Neutrophil maturation and diurnal chronomic aging represent a continuous phenotypic continuum rather than discrete cell states. NeuMapp resolves this using Spectral Diffusion Maps and Diffusion Pseudotime (DPT).
-
Tissue Compartmentalization Deconvolution: Bulk tissue biopsies, spatial arrays (10x Visium/Xenium), and organ marginal pools represent mixtures of multiple cellular niches. NeuMapp resolves the exact fractional distribution of neutrophil functional states via Constrained Non-Negative Matrix Factorization (NNLS +
$\ell_1$ sparsity) and spatial autocorrelation (Moran's$I$ ).
Deconvolves mixed spot or tissue compartment profiles
-
Spatial Autocorrelation (Moran's
$I$ ): Identifies non-random spatial clustering of specific neutrophil hubs across tissue coordinates:$$I = \frac{N}{S_0} \frac{\sum_i \sum_j w_{ij}(z_i - \bar{z})(z_j - \bar{z})}{\sum_i (z_i - \bar{z})^2}$$ - Niche Colocalization Graph: Generates a cross-correlation network quantifying spatial co-occurrence or mutual exclusion between distinct neutrophil states and surrounding stromal/immune cell types.
Reconstructs non-linear developmental and circadian aging paths along the single-cell manifold:
-
Adaptive Gaussian Transition Kernel:
$$K(x_i, x_j) = \exp\left(-\frac{|x_i - x_j|^2}{\sigma(x_i)\sigma(x_j)}\right)$$ where$\sigma(x_i)$ is determined by the distance to the$k$ -th nearest neighbor. -
Spectral Diffusion Map: Normalizes kernel into a row-stochastic Markov transition operator
$\mathbf{T}$ , computing top diffusion components (DC1, DC2, DC3) capturing global lineage topology. -
Diffusion Pseudotime (DPT): Measures random-walk geodesic distances from the bone marrow precursor root (
Pre_Neu) to terminal chronomic and activation states. - PAGA-Style Markovian Connectivity: Computes coarse-grained cluster transition probabilities between maturation stages.
Couples single-cell trajectory kinetics with spatial compartment deconvolution, mapping continuous developmental vectors directly into physical organ microenvironments.
modules/neumapp_spatial/(3,400+ lines): Histology slice overlay, spatial feature contours, spot clustering, and differential microenvironment niche exploration.modules/seurat_singlecell/& rootapp.R(1,360+ lines): Native support for Seurat v5 Assay5 multi-layer architectures, real-time dynamic subsetting, on-the-fly UMAP re-clustering, cell density glow filters, and automated differential marker detection.
# 1. Install Python dependencies
pip install -r requirements.txt
# 2. Run the end-to-end integrative pipeline (direct execution or module mode)
python3 python/pipeline_integrative.py
# or: python3 -m python.pipeline_integrative
# 3. Run standalone neutrophil spatial hub deconvolution CLI
python3 python/spatial_deconvolution.py --output-csv spatial_niche_proportions.csv
# 4. Run standalone single-cell trajectory & chronomics CLI
python3 python/trajectory_inference.py --n-neighbors 15 --output-csv trajectory_pseudotime_results.csv# 1. Install R dependencies
Rscript setup_dependencies.R
# 2. Launch root Seurat v5 single-cell explorer
Rscript -e "shiny::runApp('app.R')"Load the included demo dataset (demo_data/demo_pbmc_small.rds) directly in the UI for instant testing.
To launch the multi-module suite selector:
source("launch_suite.R")
launch_suite("spatial") # Launches NeuMapp Spatial Suite
# OR
launch_suite("singlecell") # Launches Seurat Single-Cell ExplorerTranscriptomic-Explorer/
├── launch_suite.R # Interactive multi-module launcher
├── app.R # Single-cell Shiny dashboard (root, Seurat v5)
├── setup_dependencies.R # Automated CRAN/Bioconductor dependency installer
├── pyproject.toml # PEP 621 Python packaging configuration
├── requirements.txt # Python dependencies (numpy, scipy, pandas)
├── python/ # NeuMapp Python Analytical Engine
│ ├── __init__.py # Package exports
│ ├── spatial_deconvolution.py# Constrained NNLS & Moran's I spatial niche engine
│ ├── trajectory_inference.py # Spectral diffusion maps & Markov transition graph
│ └── pipeline_integrative.py # Integrative scRNA-seq trajectory + spatial hub runner
├── modules/
│ ├── neumapp_spatial/ # Spatial Transcriptomics Shiny Suite (3400+ lines)
│ │ ├── server.R
│ │ └── ui.R
│ └── seurat_singlecell/ # Seurat v5 Manifold Explorer
│ └── app.R
├── tests/ # Automated unit and integration test suite
│ ├── test_python_modules.py # Python deconvolution & trajectory unit tests
│ └── test_r_integration.R # R architecture & Shiny launcher test suite
├── data/
│ └── sample_metadata.csv # Sample and cohort metadata schema
├── demo_data/
│ └── demo_pbmc_small.rds # Curated demo dataset for instant evaluation
├── LICENSE # MIT Open-Source License
├── CITATION.cff # Academic citation metadata
├── CONTRIBUTING.md # Developer guidelines
├── SECURITY.md # Security and vulnerability reporting
└── README.md
# Execute Python analytical test suite
python3 tests/test_python_modules.py
# Execute R architecture and module integration tests
Rscript tests/test_r_integration.RThe architectural design of NeuMapp Explorer is founded on the following landmark works on neutrophil compartmentalization, vascular hubs, and heterogeneity:
- Hidalgo, A., et al. (2019). Neutrophils: Forging the Future of Immunology. Cell, 179(3), 585–597.
- Ballesteros, I., et al. (2020). Cellular and Molecular Determinants of Neutrophil Heterogeneity and Aging across Tissues. Cell, 181(4), 842–859.
- Casanova-Acebes, M., et al. (2018). Neutrophils Instruct Homeostatic and Pathological States in Distinct Organ Niches. Cell, 174(5), 1170–1182.
- Hidalgo, A., et al. (2021). Functional Epigenomics and Vascular Niches of Leukocyte Margination. Nature Immunology, 22(8), 940–953.
Distributed under the MIT License. Copyright © 2025–2026 Maxence Tricaud.
@software{tricaud2026neumapp,
author = {Tricaud, Maxence},
title = {NeuMapp Explorer: Multi-Omic Landscape of Neutrophil Architecture, Trajectory Dynamics, and Spatial Tissue Compartmentalization},
year = {2026},
url = {https://github.com/mt93git/Transcriptomic-Explorer},
version = {2.1.0}
}