Skill: Multi-Omics Integration Strategist (ID: 204)

Overview

Designs multi-omics (transcriptomics RNA, proteomics Pro, metabolomics Met) joint analysis schemes, performs cross-validation at the pathway level, and provides systems biology-level integrated analysis strategies.

Use Cases

• Systems biology mechanism research for complex diseases
• Biomarker discovery and validation
• Drug target identification and pathway validation
• Multi-omics data quality assessment and consistency analysis

Directory Structure

.
├── SKILL.md                 # This file - Skill documentation
├── config/
│   └── pathways.json        # Pathway database configuration
├── scripts/
│   └── main.py             # Main analysis script
├── templates/
│   └── report_template.md   # Analysis report template
└── examples/
    └── sample_data/         # Sample datasets

Input

Required Files

Input Format Specifications

RNA Data (rna_data.csv)

gene_id,gene_name,log2fc,pvalue,padj,sample_A,sample_B,...
ENSG00000139618,BRCA1,1.23,0.001,0.005,12.5,13.2,...

Protein Data (pro_data.csv)

protein_id,gene_name,log2fc,pvalue,padj,sample_A,sample_B,...
P38398,BRCA1,0.85,0.002,0.008,2450,2890,...

Metabolite Data (met_data.csv)

metabolite_id,metabolite_name,kegg_id,log2fc,pvalue,padj,...
C00187,Cholesterol,C00187,-1.45,0.003,0.012,...

Integration Strategy

1. ID Mapping Layer

• RNA → Protein: Mapping through Gene Symbol / UniProt ID
• Protein → Metabolite: Association through KEGG/Reactome enzyme-reaction-metabolite
• RNA → Metabolite: Indirect association through KEGG pathway

2. Pathway Mapping

Supported databases:

• KEGG (Kyoto Encyclopedia of Genes and Genomes)
• Reactome
• WikiPathways
• GO (Gene Ontology) - Biological Process

3. Cross-Validation Methods

3.1 Directional Consistency Validation

• Whether the change direction of genes/proteins/metabolites in the same pathway is consistent
• Score: +1 (consistent), -1 (opposite), 0 (no data)

3.2 Correlation Validation

• Pearson/Spearman correlation analysis
• Cross-omics expression profile clustering

3.3 Pathway Enrichment Concordance

• Independent enrichment analysis for each omics
• Common enriched pathway identification

3.4 Network Topology Validation

• Construct cross-omics regulatory network
• Identify key nodes (Hub genes/proteins/metabolites)

Output

1. Integration Report (integration_report.md)

# Multi-Omics Integration Analysis Report