Normalized Enrichment Score (NES)

The raw, unadjusted statistic calculated by Gene Set Enrichment Analysis (GSEA). It represents the maximum deviation from zero of a running-sum statistic as the algorithm walks down a ranked gene list. A high positive ES indicates the gene set is concentrated at the top of the ranked list (e.g., upregulated), while a negative ES indicates concentration at the bottom. The ES is the foundational input for calculating the NES.

The primary correction applied to the raw Enrichment Score to produce the NES. Larger gene sets naturally accumulate higher running-sum maxima, creating a bias. Normalization divides the ES by the mean of all ES values observed for gene sets of the same size during phenotype permutation tests. This accounts for the empirical distribution of scores expected by chance for a given set size, enabling direct comparison between small, focused pathways and large, broad biological processes.

The primary statistical metric for assessing significance after NES calculation. FDR estimates the expected proportion of false positives among gene sets called significant. It is computed by comparing the observed NES distribution to the null distribution generated from phenotype permutations. An FDR cutoff of < 0.25 is standard for exploratory GSEA, while < 0.05 is used for high-confidence results. FDR corrects for the massive multiple hypothesis testing burden inherent in testing thousands of gene sets simultaneously.

A resampling strategy that preserves the complex gene-gene correlation structure of the expression dataset while destroying the phenotype-gene relationship. By randomly shuffling sample labels and recalculating enrichment thousands of times, it generates an empirical null distribution of NES values. This is preferred over gene permutation, which assumes gene independence and produces overly optimistic, inflated significance estimates. The NES is only valid when the null model accurately reflects the data's covariance.

The core group of genes within an enriched gene set that drives the enrichment signal. These genes appear at the extreme ends of the ranked list, before the running sum reaches its maximum deviation. Analyzing the leading-edge subset reveals which specific members of a broad pathway are most responsible for the observed phenotype difference. This subset is critical for identifying precise therapeutic targets and for constructing Enrichment Maps that visualize pathway crosstalk.

An alternative to GSEA that calculates pathway enrichment scores for each individual sample without requiring a defined phenotype contrast. GSVA estimates the cumulative density function of gene expression ranks within a gene set for every sample independently. While GSEA produces one NES per gene set per comparison, GSVA produces a matrix of scores, enabling downstream unsupervised learning like clustering or survival analysis. It transforms gene-level data into pathway-level activity profiles.