Hello! I am attempting to run MINT sPLS-DA on two independent data sets that I have combined. Should I run the MINT sPLS-DA directly on the combined data set, or should I transform and remove the batch effects of the independent data sets before combining them and running MINT sPLS-DA?
Hi @10RL ,
MINT should run without batch correction. However during the performance evaluation we use the ‘leave one study out’ approach. With only 2 studies you might be limited (depending on the number of samples you have per study), but worth a try.
Alternate option is to remove the batch effects first and then go with a sPLS-DA analysis without MINT.
Kim-Anh