Clarification for IntegratedLearner preprocessing

[0xMuluh]

Currently, it's unclear in the documentation that IntegratedLearner requires both overlapping samples and matching sample names across experiments.

Request

1. Clarify that the mae object must contain complete cases across experiments.
2. Explicitly state that sample names must match between experiments character-for-character.
3. Provide a method to harmonize sample names, e.g., using sampleMap or renaming with patient IDs.

This would prevent confusion and save time when preparing data for IntegratedLearner/FromMAE().

ping: himelmallick/IntegratedLearner#45