ILC Progenitors in Fetal Liver (FL) and Bone Marrow (BM)

Project Overview

This project focuses on understanding the transcriptional and functional diversity of innate lymphoid cell (ILC) progenitors across fetal liver (FL) and bone marrow (BM). It employs state-of-the-art single-cell RNA sequencing (scRNA-seq) and computational methods to:

1. Identify ILC subsets and progenitors.
2. Analyze transcriptional trajectories.
3. Integrate datasets across tissues.
4. Perform signature-based scoring to compare populations.
5. Explore regulatory networks driving ILC differentiation.
6. Map the populations defined in mouse data in human data that are publicly accessible

---

Key Analyses

1. Data Preprocessing and Integration

• Preprocessing:
  • Normalized and scaled datasets for FL, BM, and small intestine (SI).
  • Identified variable features using Seurat
  • Integrate the data using Harmony for batch correction.
• Integration:
  • Merged datasets while preserving unique tissue-specific signatures.
  • Performed dimensionality reduction (PCA, UMAP) and clustering.

2. ILC Subset Identification

• Defined marker genes for distinct ILC subsets:
  • ILC1: Tbx21, Eomes, Ifng, etc.
  • ILC2: Gata3, Bcl11b, Il4, etc.
  • ILC3: Rorc, Il22, Batf, etc.
  • ILCP: Zbtb16, Tcf7, Runx3, etc.
  • CLP: Cd34, Bcl11a, Rag1, etc.
• Visualized subset-specific signatures using DotPlots and FeaturePlots.

3. Transcriptional Trajectory Analysis

• Conducted pseudotime analysis using:
  • Diffusion maps (via destiny).
  • Monocle 3 to reconstruct developmental trajectories.
• Identified key branch points and pseudotime progression in ILC development.

4. Regulatory Network Analysis

• SCENIC Workflow:
  • Constructed gene regulatory networks.
  • Identified key transcription factors (TFs) driving differentiation.
  • Visualized regulon activity using heatmaps and t-SNE plots.
• Key regulators:
  • Eomes, Tbx21, Gata3, Runx3, Ikzf2, etc.

5. Cluster and Signature-Based Scoring

• Performed module scoring to:
  • Compare ILC1 vs. ILC3 populations.
  • Assess known signatures across clusters and tissues.
• Violin plots and heatmaps for visualizing module scores.

6. Visualization

• PCA and UMAP for dimensionality reduction.
• DotPlots and FeaturePlots for gene expression.
• Heatmaps for regulon activity and pseudotime-ordered gene expression.
• Barplots for proportional comparisons across tissues and clusters.

---

Key Results

• Identified transcriptionally distinct ILC subsets across FL and BM.
• Reconstructed differentiation trajectories highlighting transitions between progenitor states and mature ILC subsets.
• Discovered regulatory networks and transcription factors critical for ILC fate decisions.
• Demonstrated tissue-specific differences in ILC progenitor populations.

---

How to Reproduce the Analysis

1. Install Dependencies:

   • R packages: Seurat, Harmony, SCENIC, ggplot2, ComplexHeatmap, etc.
   • Ensure all required scripts and raw data are available in the repository.

2. Run Scripts:

   • Follow the order in the Scripts/ directory to reproduce each step of the analysis.

3. Generate Figures:

   • Visualization scripts automatically save outputs to the Figures/ directory.

---

Acknowledgments

For questions or issues, contact [rebuffet@ciml.univ-mrs.fr]