Source: https://github.com/aipoch/medical-research-skills

Immune Pathway Analysis

When to Use

Use this skill when the goal is to quantify immune-related pathway activity from bulk expression data and compare pathway enrichment between two sample groups.

Typical requests:

• "Run immune pathway GSVA for these samples."
• "Score immune Reactome pathways and compare case versus control."
• "Generate an immune pathway heatmap from a saved result."
• "Use a local immune gene-set table for pathway scoring."

This skill is appropriate for:

• Bulk RNA-seq or microarray-like expression matrices
• Local immune Reactome gene-set tables prepared in advance
• Two-group pathway differential analysis with limma
• Reproducible CLI execution with append-only provenance files

Execution Model

This is a hybrid skill.

1. Confirm the request is in scope with this SKILL.md.
2. Ask only for missing file paths or missing group labels.
3. Run scripts/main.R with the appropriate mode.
4. Use --mode analyze to score pathways and export tables.
5. Use --mode visualize to regenerate a heatmap from a saved result object.
6. Use --mode full to run analysis and visualization in one pass.
7. Read reference files only when you need deeper algorithm, troubleshooting, or CLI details.
8. After execution, report the output directory, scoring method, comparison groups, and the primary output files.

Completion Format

After a successful run, summarize the outcome in 3 short parts:

1. Mode and method used, plus the compared groups.
2. Output directory and key files written.
3. Important warnings that affect interpretation, such as no pathways meeting fdr_threshold and fallback to |t| ranking.

Example completion summary:

Completed immune pathway analysis in ./output/run_001 using gsva for Case versus Control. Key outputs: table/immune_pathway_diff.csv, table/immune_pathway_scores.csv, data/immune_pathway_result.rds, and plot/immune_pathway_heatmap.pdf. No pathways passed FDR <= 0.05, so the workflow used the documented fallback ranking by |t| for the top-pathway export and heatmap subset.

When to Read External Files

Situation	File to Read	Purpose
Need method details or interpretation guidance	references/algorithm.md	Review GSVA logic, limma comparison, and interpretation guidance
Need runnable commands or dependency setup	references/cli-guide.md	Reuse CLI examples, fixture notes, and validated baseline records
Need error remediation	references/troubleshooting.md	Map error codes, fallback behavior, and common fixes
Need the executable workflow	scripts/main.R	Use the CLI entry point
Need minimal demo inputs	tests/data/	Use the bundled example expression, group, and gene-set files

When Not to Use

• Immune cell fraction estimation or deconvolution
• Gene-level differential expression without pathway scoring
• Single-cell clustering, annotation, or communication analysis
• Clinical diagnosis or treatment selection

If the request falls outside these boundaries, stop and state that this skill only covers bulk immune pathway GSVA or ssGSEA analysis from a local immune gene-set table.

Input Validation

This skill accepts:

• A bulk expression matrix in CSV or TSV format
• A sample group file with exactly two comparison groups for analysis mode
• A local immune gene-set table in long format
• An existing output directory containing data/immune_pathway_result.rds for visualize mode

If the user's request does not involve bulk immune pathway scoring from local files, do not proceed with the workflow. Instead respond:

"Immune Pathway Analysis is designed for bulk immune pathway GSVA or ssGSEA analysis from a local gene-set table. Your request appears to be outside this scope. Please provide a bulk expression matrix, a two-group sample file, and a local pathway table, or use a more appropriate skill for deconvolution, differential expression, or single-cell analysis."

If the request is in scope but required inputs are missing, ask only for the missing file paths or group labels before running scripts/main.R.

Usage

Rscript scripts/main.R \
  --mode full \
  --input_file ./expression_matrix.csv \
  --group_file ./group_info.csv \
  --geneset_file ./immune_genesets.csv \
  --case_group Case \
  --control_group Control \
  --output_dir ./output/run_001 \
  --seed 42

Arguments

Short	Long	Type	Default	Description
-m	--mode	character	analyze	Run mode: analyze, visualize, or full
-i	--input_file	character	required for analyze or full	Expression matrix file in CSV or TSV format
-g	--group_file	character	required for analyze or full	Sample group file in CSV or TSV format
	--geneset_file	character	required for analyze or full	Local immune gene-set table in long format
	--geneset_column	character	gs_name	Pathway column in the gene-set table
	--gene_column	character	gene_symbol	Gene symbol column in the gene-set table
	--focus_genesets	character	optional	Comma-separated pathway names to prioritize in the heatmap
-a	--case_group	character	required for analyze or full	Case group label
-c	--control_group	character	required for analyze or full	Control group label
-o	--output_dir	character	./output	Output directory inside this skill folder
	--method	character	gsva	Scoring method: gsva or ssgsea
	--kcdf	character	Gaussian	GSVA kernel: Gaussian, Poisson, or none
	--min_sz	integer	2	Minimum gene-set size
	--max_sz	integer	5000	Maximum gene-set size
	--parallel_sz	integer	1	Worker count passed to GSVA::gsva
	--mx_diff	logical	TRUE	GSVA mx.diff flag
	--tau	double	1	GSVA tau value
	--fdr_threshold	double	0.05	FDR threshold for significance summaries
	--top_n	integer	20	Maximum number of pathways exported to the top-score matrix
	--seed	integer	42	Random seed
	--timeout_seconds	integer	0	Optional timeout in seconds; 0 disables timeout
	--plot_file	character	immune_pathway_heatmap.pdf	Heatmap file name stored under plot/
	--plot_title	character	Immune Pathway GSVA Heatmap	Heatmap title
	--width	double	14	Heatmap width in inches
	--height	double	8	Heatmap height in inches
	--colors	character	#91bfdb,#ffffbf,#fc8d59	Comma-separated heatmap colors
	--scale	character	none	Heatmap scale mode: none, row, or column
	--cluster_rows	logical	TRUE	Cluster heatmap rows
	--cluster_cols	logical	FALSE	Cluster heatmap columns
	--show_rownames	logical	TRUE	Show pathway names on the heatmap
	--show_colnames	logical	FALSE	Show sample names on the heatmap
	--fontsize	double	10	Base heatmap font size
	--fontsize_row	double	8	Heatmap row font size
	--fontsize_col	double	9	Heatmap column font size
	--legend_cex	double	1	Legend text scaling factor
	--top_up	integer	optional	Number of up-regulated pathways kept for plotting
	--top_down	integer	optional	Number of down-regulated pathways kept for plotting
	--top_mode	character	both	Heatmap subset mode: both, up, down, or total
	--sort_by	character	FDR	Pathway ranking: FDR, absLFC, or LFC
	--append_stats	logical	FALSE	Append FDR and logFC to heatmap labels
	--label_max_chars	integer	90	Maximum heatmap label length

Input Format

Expression Matrix

• CSV or TSV file
• The first column contains gene identifiers
• Remaining columns are sample names
• Values must be numeric
• Missing values are not allowed

Group File

• CSV or TSV file with a header
• Supported sample column names: sample, sample_name, sample_id, sampleid
• Supported group column names: group, condition, class, cluster
• Sample names must match the expression matrix columns

Gene-Set Table

• CSV or TSV file in long format
• One row per gene-to-pathway mapping
• Must contain a pathway column and a gene column
• Default column names are gs_name and gene_symbol
• Alternate schemas are supported through --geneset_column and --gene_column

Output Files

File	Description
table/immune_pathway_diff.csv	Differential pathway results from limma
table/immune_pathway_scores.csv	Full GSVA or ssGSEA score matrix
table/immune_pathway_scores_top.csv	Top pathway score matrix selected from the differential results
table/immune_gene_set_summary.csv	Per-pathway gene counts after table parsing
data/immune_pathway_result.rds	Saved analysis object used by visualize mode
plot/immune_pathway_heatmap.pdf	Heatmap PDF generated in visualize or full mode
session_info.txt	R session and package version record
output_manifest.txt	Append-only output manifest
run_record.txt	Append-only run record

When no pathways pass the selected fdr_threshold, the workflow logs a warning and falls back to ranking pathways by |t|. In that case, table/immune_pathway_scores_top.csv can still be populated for downstream plotting and review.

Error Handling

Error Code	Meaning	Solution
SKILL_FILE_NOT_FOUND	An input file or saved result object does not exist	Check the path and rerun
SKILL_MISSING_COLUMNS	The group file or gene-set table lacks required columns	Rename the columns or export the correct table
SKILL_EMPTY_DATA	The matrix, gene-set list, or plotting matrix is empty	Check the input content, gene overlap, and selected gene-set columns
SKILL_INVALID_PARAMETER	A CLI value is missing, invalid, or unsafe	Review the argument table and rerun
SKILL_SAMPLE_MISMATCH	Samples do not align between the matrix and group file	Align sample names before rerunning
SKILL_PACKAGE_NOT_FOUND	Required R packages are missing	Install the packages listed in references/cli-guide.md
SKILL_VERSION_INCOMPATIBLE	Installed package versions are a known incompatible combination	Follow the version guidance in references/cli-guide.md and rerun

Read references/troubleshooting.md if the error persists.

Testing

• Minimal runnable files are bundled in tests/data/.
• The bundled smoke-test dataset is intended to validate execution and the fallback path. It may legitimately produce zero pathways with FDR <= 0.05.
• tests/data/immune_genesets_minimal.csv is a unit-test fixture, not a drop-in full-workflow demo with tests/data/expression_matrix.csv unless you prepare a matching matrix with overlapping genes.
• Use Rscript tests/run_unit_tests.R to run boundary checks, helper-function checks, and validation tests without the full GSVA workflow.
• Use Rscript tests/run_tests.R to execute the unit checks plus the full smoke-test workflow.
• Use Rscript tests/test_skill.R tests/output to validate the expected outputs.
• The validated test baseline, package versions, fixture notes, and custom-column CLI examples are documented in references/cli-guide.md.