纯生信的文章还能发高分期刊吗？

# CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

## Project Overview

This is a bioinformatics project focused on rheumatoid arthritis (RA) gene expression analysis. The goal is to identify potential disease genes through differential expression analysis and weighted gene co-expression network analysis (WGCNA), ultimately generating five diagnostic figures as shown in Figure 1 of the accompanying research paper.

## Data Sources and Structure

### GEO Datasets
- **GSE205962**: Bulk RNA-seq data (GPL16043 platform)
- **GSE100191**: Bulk RNA-seq data (GPL13497 platform)
- **GSE17755**: Validation dataset using peripheral blood samples (GPL1291 platform)
- **GSE159117**: Single-cell RNA-seq data for scRNA-seq analysis

### Directory Structure

data/
├── GSE100191/          # Bulk RNA-seq expression data
├── GSE17755/           # Validation dataset (peripheral blood)
├── GSE159117/          # Single-cell RNA-seq data
└── GSE205962/          # Primary bulk RNA-seq dataset

workflow/               # Processing scripts (currently empty - needs population)
result/                 # Output figures and analysis results (currently empty)


## Analysis Workflow

The project follows a multi-step bioinformatics pipeline:

1. **Data Preprocessing**
   - Batch effect removal using ComBat() function from sva package
   - Probe annotation mapping (GPL16043, GPL13497, GPL1219 platforms)
   - Gene symbol ID assignment with maximum expression retention for multiple probes

2. **Differential Expression Analysis**
   - Uses limma package for RA vs healthy control (HC) comparison
   - Thresholds: p < 0.05, |log2 FC| ≥ 0.585
   - Fold change calculation as ratio of disease to normal expression means

3. **WGCNA Analysis**
   - Scale-free co-expression network construction
   - Module identification and consolidation
   - Gene significance (GS) and module membership (MM) calculation
   - Thresholds: MM ≥ 0.6, GS ≥ 0.1

4. **Integration**
   - Merge WGCNA and DEGs results to identify potential disease genes

## Expected Outputs

Generate five diagnostic figures (Figure 1):
- A: Volcano plot of differential expression analysis
- B: Heatmap of differential expression analysis
- C: Soft threshold selection and cluster dendrogram
- D: WGCNA correlation heatmap (modules vs HC/RA traits)
- E: Inter-module correlation heatmap with disease trait associations

## Technical Requirements

### R Environment
- R version 4.3.3
- Required packages:
  - sva (ComBat functionfor batch effect removal)
  - limma (differential expression analysis)
  - WGCNA (weighted gene co-expression network analysis)

### Data Processing Notes
- Multiple microarray probes mapping to same gene: retain maximum expression value
- Statistical methods: Spearman's rank correlation, Mantle's test
- Expression filtering: genes with average FPKM > mean threshold

## Development Guidelines

- **Reference Material**: Use article.pdf as the authoritative sourcefor methodology when implementation details are unclear
- **Script Location**: Place all final processing scripts in the `workflow/` directory
- **Output Location**: Save all generated figures and results in the `result/` directory
- **Language**: Comments and documentation should be in Chinese as specified in the original requirements
- **Intermediate Files**: Do not save intermediate processing files or script revisions - only keep final working versions

## Data File Formats

- **CEL files**: Affymetrix microarray data (.CEL.gz)
- **Series Matrix**: GEO series expression matrices (.txt.gz)
- **Platform Files**: Probe annotation files (.txt, .annot)
- **Single-cell**: 10X Genomics H5 format for scRNA-seq data