16 months ago by
Spain/Barcelona/Universitat Pompeu Fabra
We did not develop the ssGSEA method, we just implemented it in the GSVA package. The methodological descriptions in the online manual and the vignette of GSVA refer to the GSVA (deafult) method. The ssGSEA method was introduced by Barbie et al. (2009), where you can find its description in the Online Methods, subsection 'Signature projection method'. That said, GSVA and ssGSEA use both a competitive null hypothesis, meaning that, for each gene set, they compare genes inside the gene set against genes outside the gene set, along a ranking of genes on the basis of their expression. A enrichment score of zero implies that the distribution of the genes inside the gene set along the ranking does not differ from the distribution of genes outside the ranking.