Run your first analysis: Riaz ON

This example uses melanoma biopsies collected during nivolumab treatment in Riaz et al. (2017). It compares complete or partial responders (CR/PR) with patients with progressive disease (PD). Positive values are higher in responders.

Run the example

Install lisaR first, then copy the following block into R or RStudio. It prepares the human gene-set resource, downloads the previously computed DE table and runs one STANDARD analysis. It does not fit DESeq2 again or run PRE and ON together.

Use a short writable path. The riaz-data and riaz-on directories must not already exist; choose another root if you have run this example before. The resource is downloaded once and reused on later runs. The accept_terms = TRUE argument records acceptance of the applicable resource terms.

library(lisaR)

# Use a short, writable path. Change C:/lisa if needed.
root <- if (.Platform$OS.type == "windows") "C:/lisa" else "~/lisa"
root <- path.expand(root)
dir.create(root, recursive = TRUE, showWarnings = FALSE)

# Prepare the human gene-set membership resource once.
prepare_lisa_msigdb_resource(accept_terms = TRUE)

# Download the prepared Riaz data to a new directory.
riaz <- install_lisa_example_bundle(
  "riaz-gse91061",
  destination = file.path(root, "riaz-data")
)

# Analyse ON only: CR/PR responders versus PD during treatment.
result <- run_lisa(
  de = list(responders_vs_pd_on = file.path(
    riaz$bundle, "outputs", "lisa_inputs", "responders_vs_pd_on.tsv"
  )),
  output_dir = file.path(root, "riaz-on"),
  species = "Homo sapiens",
  method = "DESeq2",
  positive_direction =
    "Higher in CR/PR responders than in PD during treatment",
  mode = "standard"
)

stopifnot(identical(result$gate, "PASS"))
browseURL(result$report_index)

The script is also installed at system.file("examples", "riaz-on-quick-start.R", package = "lisaR").

What to expect

The report opens when the run finishes. Enrichment and figure generation can take time; this is a real-data analysis, not an instant demonstration. If R reports an error, resolve it before opening a partial report.

Open responders_vs_pd_on, then its GOBP-C2 LISA plot. Select a category to inspect its member gene sets. PNG and SVG figures, source tables and figure recipes are included. PDF is optional.

Keep the whole riaz-on folder if you move or share the report, not just report_index.html.

Next: Use your own DE. To explore the example output first, read your report.