npGSEA

npGSEA is a computational method designed to expedite the analysis of gene set enrichment in high-throughput gene expression data, explicitly addressing the challenge of testing the association between collections of genes and outcomes of interest. Traditional permutation-based gene set tests require many permutations to achieve statistically significant p-values, especially when evaluating many gene sets, making the process computationally intensive and time-consuming. npGSEA overcomes this limitation by employing a moment-based parametric approximation to the permutation distributions for gene set tests, significantly reducing the computational burden.

The method focuses on two types of gene set statistics: sums and sums of squared correlations, identified as top performers in comprehensive simulations. By calculating the exact relevant moments of these statistics and fitting them to parametric distributions, npGSEA achieves a computational efficiency comparable to conducting a fraction of the permutations traditionally required. This efficiency gain is particularly notable for quadratic statistics, where the computational cost is reduced to orders of magnitude less than standard permutation sampling.