Question 1

What data does the Pathway Analysis pipeline accept?

Accepted Answer

A DESeq2-based differential gene expression table for each comparison you want to analyze. That table normally comes from a DESeq2 run on bulk RNA-seq gene counts, but any DGE result with gene identifiers, fold-changes and adjusted p-values works. If your data is still raw reads, the chain is Quality Control, then Bulk RNA-seq, then a DESeq2 run, and then this pipeline.

Question 2

What comes back from a Pathway Analysis run?

Accepted Answer

Per comparison: GO, KEGG and Reactome enrichment tables, GSEA results on the fold-change-ranked gene list, enrichment plots for the top pathways, and one HTML report that brings all the plots and tables together. Everything lands in the Pipeline Results folder, so the raw tables and the static plot images sit side by side.

Question 3

How do I find the plots and tables after a run?

Accepted Answer

Open the Pipeline Results folder for the run and search it. The static images for the KEGG and GSEA analyses are stored as plot files alongside the raw result tables, so you can pull a figure straight into a slide or pull the underlying table into your own script. The HTML report links to the same files, so you can also browse first and download what you need.

Question 4

How do I read the enrichment tables?

Accepted Answer

Each table lists pathways or GO terms with the genes from your list that fall in them, and an adjusted p-value for the enrichment. Start with the terms that are significant in the over-representation tables and also move in the same direction in the GSEA results: agreement between the two methods is the strongest signal. Terms that appear in both KEGG and Reactome are the ones to trust first.

Question 5

What is the difference between over-representation and GSEA?

Accepted Answer

Over-representation asks whether the significantly up- or down-regulated genes (adjusted p-value below 0.05) contain more of a pathway's genes than chance would predict. GSEA uses the full ranked list of all genes by fold change, so it can catch pathways where many genes shift modestly in the same direction without any single gene crossing the significance line. The pipeline runs both because they answer slightly different questions.

Question 6

Which pathway databases are used?

Accepted Answer

Three. GO biological processes, tested by over-representation and GSEA. KEGG pathways, tested the same way. And Reactome, tested by over-representation on the up and down gene lists. Running all three means a pathway that matters to your experiment is unlikely to be missed just because it is curated in one database and not another.

Question 7

What should I have ready before starting a run?

Accepted Answer

One DGE result table per comparison, with the comparisons named clearly, and agreement with your bioinformatician on which comparisons matter. The pipeline prepares the gene lists itself: significant up- and down-regulated genes at an adjusted p-value below 0.05, plus the full fold-change-ranked list for GSEA. If you are unsure the DGE table is in the right shape, the DESeq2 guide shows what a clean result looks like.

Pathway Analysis

Input

Output

From gene lists to pathways

Why each step is here

Input preparation

GO enrichment

Enrichment plots

KEGG analysis

Reactome analysis

HTML report

What comes back

Pathway Analysis FAQ

Take the next step

Run it in Horizon

Need a tailored pipeline?