Spatial Virtual Gene Knockout: Code and Data for Atherosclerotic Plaque Spatial Transcriptomics Analysis

This package contains all analysis code and processed data outputs for the study "Spatial Virtual Gene Knockout: Identifying Disease-Critical Network Nodes in Atherosclerotic Plaque Transcriptomics." The Spatial VGK method extends Virtual Gene Knockout (VGK) to spatial transcriptomics data, enabling spatially-aware identification of disease-critical genes in human carotid atherosclerotic plaques. By integrating spatial coordinates with gene co-expression networks, it detects genes whose network disruption the cell type context modulates — revealing perturbation patterns invisible to traditional differential expression analysis. Contents: - code/ (44 Python scripts): Core VGK algorithm, spatial analysis, region-specific VGK, cell-type VGK, bulk vs spatial comparison, figure generation, and manuscript preparation. - data/ (27 files): Spatial VGK DI scores for 1,999 genes, network properties, region-specific results across 8 plaque regions, cell-type-specific DI matrices for 8 cell types (126,860 cells), and bulk-spatial overlap analysis. Data sources: - Spatial transcriptomics: EGA EGAD50000000936 (study EGAS50000000660) - Single-cell RNA-seq atlas: Zenodo record 14007461 - Bulk transcriptomics: GEO GSE100927 Requirements: Python 3.10+, numpy, pandas, scipy, matplotlib, seaborn, networkx, squidpy, scanpy, python-docx.