Hi, I obtained a list of genes selected with my Machine Learning algorithm, now I would like to perform an enrichment pathway analysis with GSEA preranked, using the weights associated to each gene of my list by the Machine Learning model. My weights are all positive by construction, as opposed to a standard DE analysis. The first question is: do I have to transform or normalize the weights? Secondly: do I have to provide a background list of genes? In that case the background would have to match the genes expressed in my case study? Thank for your help.