Principal Component Analysis (PCA) is a valuable method for the analysis and validation of molecular dynamics simulation data. Metrics extracted from PCA provide a robust check on issues such as equilibration and sampling. The PCA process can highlight the most significant structural changes that occur during a simulation, and is a powerful tool for the comparative analysis of multiple datasets.
PyPcazip provides a suite of command line tools for PCA analysis of trajectory data in any of the common MD formats, and also a Python library that can be used in the researcher's own analysis scripts.

For more details see here.

MDP PCA Diagram