The paper introduces Gene Set Variation Analysis (GSVA), a novel method for gene set enrichment (GSE) analysis that estimates the variation of pathway activity across a sample population in an unsupervised manner. GSVA addresses the limitations of traditional GSE methods, which are often limited to case-control studies and may not robustly handle highly heterogeneous data sets. GSVA is demonstrated to be more powerful in detecting subtle changes in pathway activity compared to existing methods, both in simulated and real data. The method is applied to both microarray and RNA-seq data, showing its flexibility and broad applicability. GSVA is implemented as an open-source software package available in R, part of the Bioconductor project. The authors conclude that GSVA provides a robust starting point for building pathway-centric models of biology, particularly for RNA-seq data, and can be used for various downstream analyses such as differential pathway activity identification and survival prediction.The paper introduces Gene Set Variation Analysis (GSVA), a novel method for gene set enrichment (GSE) analysis that estimates the variation of pathway activity across a sample population in an unsupervised manner. GSVA addresses the limitations of traditional GSE methods, which are often limited to case-control studies and may not robustly handle highly heterogeneous data sets. GSVA is demonstrated to be more powerful in detecting subtle changes in pathway activity compared to existing methods, both in simulated and real data. The method is applied to both microarray and RNA-seq data, showing its flexibility and broad applicability. GSVA is implemented as an open-source software package available in R, part of the Bioconductor project. The authors conclude that GSVA provides a robust starting point for building pathway-centric models of biology, particularly for RNA-seq data, and can be used for various downstream analyses such as differential pathway activity identification and survival prediction.