I have a single cell dataset which I consider to be a reference , let say, for an human organ, I have identified some new clusters that correspond to cell types. I also have WGCNA modules that correspond to a identified cluster. Given this background, is there any possibilities to validate the presence of the cell types from my reference dataset inside another single cell dataset ?
I was thinking about GSEA, but all the gene in a set have the same importance. Could I use eigen vectors from first principal component for a given module expressed in the corresponding cell type, as a weight. Then calculate inside another dataset a sort of "first principal component" using the previous eigen vectors. Then I would simply calculate the correlation between the first principal component and the calculated one to validate whether my cell type is present or not. I was also thinking about a sort of weigthed GSEA, if anything like this ever existed.