PCA as a Practical Indicator of OPLS-DA Model Reliability

Abstract

Supported by the results of Monte Carlo analyses of PCA group separations and OPLS-DA cross-validation metrics, we provide practical guidelines and cross-validatory recommendations for reliable inference from PCA and OPLS-DA models.