Overview

• Estimate of total amount of time this assignment is going to take⏰:

Day 1 - March 31st, 2025

Estimate time to complete the following tasks: 3 hours - ended up working 3.5 hours not finishing the tasks

Tasks

• create ranked list ✅
• run and analyze gsea ✅
• get first cytoscape viewing - not done

Workflow

docker run -e PASSWORD=changeit  -v "$(pwd)":/home/rstudio/projects -p 8787:8787 risserlin/bcb420-base-image:winter2025-arm64

• Got feedback from TA to reduce the coverage of the previous assignment and hide some of the code. Found out that you can hide the contents and outputs of a code block with include = FALSE in the code chunk header.
• Initially, I couldn't get GSEA to run, it would keep on erroring. At first I thought it might be an issue with the gmt file because that is what I struggled with in the GSEA assignment but after inspecting my ranked list, I realized that the gene names were omitted. I fixed my R code to change this. (see below)

# before: did not include gene names
ranked_list <- 
  -log10(qlf_output_hits_table$FDR) * sign(qlf_output_hits_table$logFC)
names(ranked_list) <- rownames(ranked_list)
ranked_list <- sort(ranked_list, decreasing = TRUE)

# after: includes gene names
ranked_list <- 
  -log10(qlf_output_hits_table$FDR) * sign(qlf_output_hits_table$logFC)
names(ranked_list) <- rownames(qlf_output_hits_table)
ranked_list <- sort(ranked_list, decreasing = TRUE)

EnrichmentMap in Cytoscape

Writing out the steps I took because they were not obvious - I had to rewatch all the lectures as my notes were not step by step.

• Installed Cytoscape v3.10.3 for MacOS, EnrichmentMap v3.5.0 from the website.
• Opened the Cytoscape app, opened EnrichmentMap from the "Apps" section in the menu at the top.
• Opened the data input panel.
• Input and parameters as shown in the screenshot of the EnrichmentMap input page below.
• Selecting cluster cmd + click + drag
• In the "Style" panel, I set the mapping variable to "NES", and it was very apparent that the majority of the gene sets were down enriched (there was only one visible node that had a strong, positive NES value). A screenshot of this initial network is below.
• The single positively enriched node was the "HALLMARK_TNFA_SIGNALING_VIANFKB"

Day 2 - April 1st, 2025

Estimate time to complete the following tasks: 8 hours - worked 2 hours, could not continue due to sickness

Tasks

• get first cytoscape viewing ✅
• create themes (groups) ✅
• analyze one pathway (probably the one in the largest grouping) - not done
• write report - not done

Workflow - continued from yesterday

• Installed AutoAnnotate app v1.5.2
• Did the annotations with the default parameters but the labels overlapped so I ran it a second time, looks good now
• Installed yFiles Layout Algorithms app v1.1.5

Day 3/4 - April 2nd-3rd, 2025

Mostly sick, slow progress.

Tasks

• analyze one pathway (probably the one in the largest grouping) ✅

Workflow

• noticed smaller clusters are from specific data bases (ex GO or REACTOME)

• the largest theme DNA strand repair has multiple, might be interesting to look at hub pathways (nodes connecting bigger clusters)

• tried REACTOME but Wiki pathways was easier to use as it was integrated in cytoscape

• WikiPathways is app but is accessed by "importing data" (below)

• selected "DNA REPAIR PATHWAYS FULL NETWORK" as gene set

Day 5/6 - April 4-5th, 2025

Recovering from bad health. Expected time : 4 hours - Actual time: 8 hours

Tasks

• do full analysis ✅
• write up rest of the report, figure legends etc ✅

Notes

• the individual nodes left out of clusters/themes were simply those that did not share genes with other genesets
• had to reread the original paper & do literature review to analyze (this took most of my time) - notes in Notion