scale the data based on each individual paired group?

[Tonny-zhou]

Dear author,

Thank you very much for developing this convenient metabolomics analysis tool. I have a small question that I would like to ask you. Upon reviewing the source code, I noticed that in the runPLSDA function, the metabolomics data is scaled (default: uv). However, it seems that the scaling is applied to all the grouped data together. For example, if I have two paired groups (A vs B; A vs C), the scaling is performed on all the data from groups A, B, and C combined. Might it be more appropriate to scale the data based on each individual paired group?

Thank you for your attention and I look forward to your insights on this matter.

Best regards,
Tonny

[wenbostar]

The scaling is typically applied to the whole dataset. You could try to perform the scaling on each group separately to perform PLSDA analysis to see if there is any obvious difference when comparing to the scaling on the whole dataset.

To do the scaling on each group, you could prepare two datasets, one contains A and B and the other one contains A and C.