How yyyuechen/ai-omics-for-cancer-regulatory-gene-identification works
This repository is a reproduction workflow for finding and interpreting candidate regulatory genes in aflatoxin B1-induced hepatocellular carcinoma. Its main path runs from GEO expression preprocessing and target collection through differential expression, co-expression and network analysis, then machine-learning selection and signature validation. Molecular docking is a distinct analysis branch. The README documents these stages and their order; source excerpts were unavailable, so links represent documented workflow handoffs rather than verified function calls.
Main entry points: 05_PPI_and_enrichment_analysis.R PPI and enrichment · 01_HCCdata_preparation.R Expression preprocessing · 03_batch_correction_and_DEG_analysis.R Batch correction and DEGs · 04_WGCNA_module_analysis_and_DEG_integration.R WGCNA and DEG integration
Components 15
Transcriptomics
- Expression preprocessing
scripts/01_HCCdata_preparation.R - Batch correction and DEGs
scripts/03_batch_correction_and_DEG_analysis.R - WGCNA and DEG integration
scripts/04_WGCNA_module_analysis_and_DEG_integration.R
Candidate Networks
- Target integration
scripts/02_target_database_integration.R - PPI and enrichment
scripts/05_PPI_and_enrichment_analysis.R - Integrated candidates
scripts/05_PPI_and_enrichment_analysis.R
Machine Learning
- Cohort preparation and Lasso-RF
scripts/06_data_preparation_and_LassoRF.R - Model comparison
scripts/07_machine_learning_model_comparison.R - Signature validation and interpretation
scripts/08_final_signature_model_validation_and_interpretation.R
Molecular Docking
- Docking workflow
scripts/09_Molecular_docking_analysis.txt
Other components
- Researcher
- GEO datasets
- Target resources
- Structure and ligand resources
- Docking and visualization tools
Connections 21
- Researcher → Expression preprocessing: runs
- GEO datasets → Expression preprocessing: provides datasets
- Expression preprocessing → Batch correction and DEGs: provides expression data
- Researcher → Target integration: runs
- Target resources → Target integration: provides target candidates
- Researcher → Batch correction and DEGs: runs
- Batch correction and DEGs → WGCNA and DEG integration: provides DEG results
- Researcher → WGCNA and DEG integration: runs
- WGCNA and DEG integration → PPI and enrichment: provides integrated candidates
- Target integration → PPI and enrichment: provides AFB1 candidates
- Researcher → PPI and enrichment: runs
- PPI and enrichment → Integrated candidates: produces
- Integrated candidates → Cohort preparation and Lasso-RF: provides candidate genes
- Researcher → Cohort preparation and Lasso-RF: runs
- Cohort preparation and Lasso-RF → Model comparison: provides cohort data
- Researcher → Model comparison: runs
- Model comparison → Signature validation and interpretation: informs signature validation
- Researcher → Signature validation and interpretation: runs
- Researcher → Docking workflow: follows tutorial
- Structure and ligand resources → Docking workflow: provides structures and ligand
- Docking workflow → Docking and visualization tools: uses