Hello,
I’m going to be performing P-integration soon and I was looking at MINT (s)PLS. Should a different approach be taken if the P-variables do not perfectly overlap between datasets? For example, what should be done if the list of measured metabolites in one study is similar to but slightly different from that in another study? Thanks!
Nathan
Hi @Nathan,
P-integration needs to have the same variables across the different datasets, if you have a few missing values in your dataset here and there this can be accounted for in mixOmics models (see here), but if a whole variable or sample is missing from one dataset it would have to be removed prior to integration. If there is a set of metabolites that overlap between your two studies I would suggest just using those in MINT (s)PLS.
Hope that helps!
Eva