Immune Checkpoint Lectin Ranking in PDAC Single-Cell Analysis

Overview

This repository contains a computational framework for identifying and ranking lectins related to immune checkpoint pathways in pancreatic ductal adenocarcinoma (PDAC) using single-cell multi-modal data analysis. The analysis integrates glycan binding patterns with RNA expression profiles to discover cell-type-specific lectin interactions that may play crucial roles in immune checkpoint regulation.

Research Objective

The primary goal is to identify lectins that show specific binding patterns to immune cells in the PDAC tumor microenvironment, which could serve as: - Novel immune checkpoint targets - Biomarkers for immune cell subtypes - Therapeutic targets for cancer immunotherapy

Methodology

Multi-Modal Data Integration

• RNA Expression Data: Single-cell RNA sequencing data capturing receptor gene expression
• Glycan Binding Data: Lectin microarray data measuring glycan-lectin interactions
• Cell Types Analyzed:
  • Cancer cells (Classical, Basal-like, Intermediate)
  • Immune cells (T cells, TAMs, MDSCs, dendritic cells, B cells, etc.)

Scoring Algorithm

The analysis employs a sophisticated scoring system that evaluates: 1. RNA Specificity Score: How specifically a lectin receptor is expressed in immune cells 2. RNA Expression Level: The magnitude of receptor expression in immune cells 3. Glycan Specificity Score: How specifically a lectin binds to cancer cells 4. Glycan Binding Level: The strength of lectin binding to cancer cells

Key Features

• Emphasizes single cell-type specificity using coefficient of variation and ratio metrics
• Customizable weighting system for different scoring components
• Identifies top-expressing cell types for both RNA and glycan data

Installation

bash conda create -n scglycan python=3.12 conda install -y -n scglycan -c conda-forge r-essentials r-base r-seurat r-pheatmap r-patchwork r-ggplotify r-languageserver conda activate scglycan

Usage

1. Ensure the PDAC single-cell dataset (pdac_ctype.RData) is placed in the data/ directory

[!NOTE] The pdac_ctype.RData file should be requested from the corresponding author (Dr. Hiroaki Tateno: h-tateno[at]aist.go.jp)

1. Run the main analysis: r # In R or RStudio quarto::quarto_render("immune_checkpoint_lectin_ranking.qmd")
2. Results will be saved to:
   • data/glycan_ranking.csv: Ranked list of lectins with scores
   • data/glycan_ranking_top10.png: Visualization of top-ranked lectins

Repository Structure

├── immune_checkpoint_lectin_ranking.qmd # Main analysis notebook ├── immune_checkpoint_lectin_ranking.R # Generated R script ├── scripts/ │ └── load.R # Data exploration script └── data/ ├── pdac_ctype.RData # Input: Seurat object with multi-modal data └── glycan_ranking.csv # Output: Lectin rankings

Biological Significance

This analysis provides insights into: - Lectin-mediated immune checkpoint mechanisms in PDAC - Cell-type-specific glycosylation patterns in the tumor microenvironment - Potential therapeutic targets for enhancing anti-tumor immunity

License

This project is licensed under the MIT License - see the LICENSE file for details.