pathway-enrichment

Pathway Enrichment - Pathway Enrichment Analysis Skill

Skill Overview

The Pathway Enrichment skill provides comprehensive pathway and gene-set enrichment analysis for gene lists and ranked gene data, helping you understand the biological significance behind differentially expressed genes, CRISPR screening hits, or clustered marker genes.

Use Cases

1. Functional interpretation after differential expression analysis

After completing differential expression analysis such as DESeq2, edgeR, or limma, use this skill to identify which biological pathways, GO terms, or gene sets are enriched among significantly changed genes. Suitable for various multi-omics studies including transcriptomics and proteomics.

2. Biological interpretation of gene screening results

When you have a gene hit list (e.g., CRISPR screen hit genes, ChIP-seq peak-associated genes, or single-cell cluster marker genes), this skill helps you discover the biological processes and signaling pathways shared by these genes.

3. Pathway activity scoring for single-cell and spatial transcriptomics

Using ssGSEA or GSVA, compute pathway activity scores for each sample or cell to reveal pathway changes associated with cellular heterogeneity, developmental trajectories, or disease states.

Core Functions

1. Dual-method enrichment analysis (ORA and GSEA)

Supports over-representation analysis (ORA) based on the hypergeometric test and GSEA analysis based on ranked gene sets. The appropriate method is automatically recommended according to your data type. ORA is suitable for a clearly defined gene list, while GSEA is suitable for a complete ranked gene list with statistics, enabling detection of broader and subtler biological signals.

2. Multi-database integration and gene ID mapping

Integrates major databases such as MSigDB Hallmark, GO Biological Process, KEGG, Reactome, and WikiPathways, and performs automated analysis using gseapy and g:Profiler. Includes built-in gene ID conversion to support mapping from multiple ID formats such as Ensembl and Entrez to gene symbols.

3. Visualization and publication-ready result output

Offers a variety of visualization options, including scatter plots, bar charts, enrichment maps, and GSEA running score plots. Automatically generates publication-required statistics such as adjusted p-values, FDR, and gene overlap information. Also supports removing redundant GO terms to ensure interpretability and suitability for publication.

Frequently Asked Questions

How should I choose between ORA and GSEA?

The choice mainly depends on your data type. If you have a clear gene list (e.g., differentially expressed genes with padj < 0.05, or CRISPR screening hit genes), use ORA over-representation analysis. If you have a complete ranked gene list (each gene has a statistic or score), use GSEA or Preranked GSEA. GSEA is better for detecting widespread and subtle effects and does not require manually setting thresholds.

Why does my enrichment analysis yield no significant results?

The most common reasons include: gene ID format mismatch (e.g., using Ensembl IDs when the database expects gene symbols), incorrect background set settings (using the whole genome instead of the actually detected gene set), gene lists that are too short (fewer than 10 genes) or too long (more than 2000 genes leading to reduced specificity). It is recommended to check gene ID formats, species settings, and the definition of the background set.

How do I handle redundant terms in GO enrichment results?

The GO database contains many overlapping terms. Reporting all significant results directly can introduce redundancy. It is recommended to use an enrichment map (clustering based on term similarity), leading-edge gene overlap analysis, or report only parent terms. This skill provides enrichment map functionality based on gseapy, which can automatically identify and merge highly redundant terms, helping you select representative biological processes for reporting.