In the past few months, I have been running PLS-DA permutation tests, but the resulting P-values are always incorrect. Although the three groups are clearly separated, the permutation P-value for PLS-DA is close to 1, which is unreasonable. To verify this potential issue, I selected two groups with significant differences and performed both OPLS-DA and PLS-DA. The OPLS-DA model showed a complete separation with a P-value of 0.002, whereas the PLS-DA model also separated the groups but yielded a P-value of 1. Why is this happening?
What can I do?Can you help me ,please.