Computational analysis of a conserved immune-cell recruitment program across transplantation, infection and inflammatory disease.
This repository contains the analysis code accompanying the ALARM study:
Sinha D, Laurent T, Broquet A, et al.
A gene-expression module identifies circulating immune cells with enhanced recruitment to sites of inflammation.
iScience. 2026;29(1):114227.
DOI: 10.1016/j.isci.2025.114227
Circulating immune cells must transition from blood into inflamed tissues, but the transcriptional programs associated with this recruitment are not fully understood. Using longitudinal single-cell transcriptomics from kidney transplant recipients, the study identified a conserved gene-expression module termed ALARM, enriched for transcriptional regulators, homing receptors and early activation markers.
ALARM-high immune cells were reduced in peripheral blood during rejection while being enriched in rejecting graft tissue and in a large-animal transplantation model, consistent with preferential recruitment to inflammatory sites. The module includes CXCR4, and functional experiments showed that CXCR4–CXCL12 signalling promotes T-cell migration, early activation and a metabolic shift toward glycolysis.
Across external datasets, ALARM was also associated with immune-cell redistribution during infection and showed predictive value across multiple immune-mediated diseases, with additional validation in a pneumonia mouse model.
The repository captures several complementary layers of analysis:
- longitudinal single-cell RNA-seq analysis;
- unsupervised gene-program discovery with cNMF;
- cell-type and module-usage association analyses;
- cross-species validation;
- bulk-transcriptomic validation;
- migration-related single-cell analysis;
- ligand–receptor / CXCR4–CXCL12 biology;
- metabolic modelling with COMPASS;
- disease classification and predictive modelling.
The manuscript code is organised largely by figure and analytical objective:
ALARM/
└── Manuscript/
└── codes/
├── figure2_mod_discovery/
├── figure3_pig/
├── figure4_bulk/
├── figure6_transwell_scRNA/
├── figure7_transwell_compass/
└── figure10_classification/
| Directory | Main purpose |
|---|---|
figure2_mod_discovery |
ALARM module discovery and cell-type association; includes cNMF workflow |
figure3_pig |
Cross-species validation in the pig transplantation model |
figure4_bulk |
Validation across bulk transcriptomic datasets |
figure6_transwell_scRNA |
Single-cell analysis of migration / transwell experiments |
figure7_transwell_compass |
Metabolic-state analysis using COMPASS |
figure10_classification |
Disease classification and predictive analyses |
For example, the module-discovery workflow includes data preparation, cNMF execution, definition of module genes and cell-type/module-usage analyses.
The main result is not simply a disease-associated expression signature. ALARM represents a shared immune-cell state associated with recruitment from circulation into inflamed tissue, connecting transcriptional regulation, chemotaxis, activation and cellular metabolism.
This makes the project an example of computational immunology moving from single-cell pattern discovery through cross-cohort and cross-species validation to experimentally supported biological interpretation.
The final peer-reviewed article is available from iScience:
- DOI: https://doi.org/10.1016/j.isci.2025.114227
- PubMed: https://pubmed.ncbi.nlm.nih.gov/41531734/
- PMC: https://pmc.ncbi.nlm.nih.gov/articles/PMC12794507/
If you use this code or build on the ALARM framework, please cite:
Sinha D, Laurent T, Broquet A, et al. A gene-expression module identifies circulating immune cells with enhanced recruitment to sites of inflammation. iScience. 2026;29(1):114227. https://doi.org/10.1016/j.isci.2025.114227
Debajyoti Sinha
Nantes Université · CHU Nantes · Inserm
Computational biology · single-cell transcriptomics · computational immunology