6 GO-Enrichment Tab

6.1 Analysis Methods

6.1.1 Over Representation Analysis (ORA)

ORA is designed for analyzing specific lists of proteins, such as: - Significantly changed proteins - Proteins from a cluster - Manually curated lists of interest

Key features: - Uses Fisher’s exact test for statistical analysis - Compares your protein list against all known proteins in each biological process - Perfect for focused analysis of specific protein sets - Results show enrichment of your proteins in GO terms compared to background

6.1.2 Correlation Adjusted MEan RAnk gene set test (CAMERA)

CAMERA takes a more comprehensive approach by analyzing your entire dataset: - Uses all proteins and their measurements - Ranks proteins based on both abundance changes and statistical significance - Calculates a score using: -log10(p-value) * (Direction of the Fold-Change) - Ideal for discovering subtle but consistent changes across multiple proteins - Integrated with the Limma package for robust statistical analysis

6.2 Interface Components

6.2.1 Settings Panel

The settings panel allows you to configure your analysis:

1. Analysis Method Selection
   • Choose between ORA and CAMERA
   • Each method has specific requirements and use cases
2. GO Term Size Filters
   • Minimum term size (default: 50)
   • Maximum term size (default: 500)
   • Adjust based on your research focus:
     • Small range (10-50) for specific processes
     • Large range (250-1000) for broader processes
3. Statistical Options
   • Choose between raw p-values or adjusted p-values
   • All results undergo Benjamini-Hochberg correction

6.2.2 Results Visualization

6.2.2.1 Bar Plot

• X-axis: (adjusted) P-value
• Y-axis: Ranked GO terms
• Interactive selection of terms
• Filterable by GO categories (BP, MF, CC)

6.2.2.2 Volcano Plot

Shows different metrics based on analysis method: - ORA: Percentage of significant proteins per GO term - CAMERA: Average scoring value - X-axis: (adjusted) P-value - Y-axis: Enrichment metric

6.2.2.3 Network Plot

Visualizes relationships between GO terms: - Node size based on: * ORA: Percentage of proteins found * CAMERA: Adjusted P-value - Edge thickness shows term overlap - Interactive clustering option - Customizable edge cutoff and term count

6.2.3 Term-Specific Analysis

When selecting a specific GO term, you get access to:

6.2.3.1 Term-Protein Network

• Interactive visualization of term-protein relationships
• Multiple coloring options:
  • P-value
  • Adjusted P-value
  • Log2 Fold Change
  • -log10(adjusted P-value) * Fold Change Direction
• Shows protein connections between terms

6.2.3.2 Detailed Information

1. Heatmap
   • Z-scored log2 LFQ values
   • Hierarchical clustering of proteins
2. General Info
   • Term name and ID
   • Description
   • Statistical metrics
3. Enrichment Plot (CAMERA only)
   • Protein ranking visualization
   • Distribution of term members

6.3 Practical Tips

1. Choosing Analysis Method
   • Use ORA for specific protein lists
   • Use CAMERA for exploratory analysis
   • Consider biological context when interpreting results
2. Term Size Selection
   • Smaller ranges (10-50) for specific processes
   • Larger ranges (250-1000) for pathway analysis
   • Default (50-500) works well for most analyses
3. Result Interpretation
   • Consider both p-values and effect sizes
   • Look for biological coherence
   • Use network visualization for context
4. Data Export
   • Results can be downloaded as CSV
   • Includes all statistical metrics
   • Preserves protein annotations

6.4 References

1. Gene Ontology Consortium (http://geneontology.org/)
2. Wu D, Smyth GK. Camera: a competitive gene set test accounting for inter-gene correlation. Nucleic Acids Research. 2012
3. Alexa A, Rahnenfuhrer J. topGO: Enrichment Analysis for Gene Ontology. R package version 2.48.0, https://bioconductor.org/packages/topGO