Principal Component Analysis: A Tool to Elucidate the Essential Movements of Biomolecules

Abstract

Principal Component Analysis (PCA) is a widely used procedure for examining data collected in biomolecule simulations. Its main virtue is that it significantly reduces the dimensionality of the molecule's configurational space. In simple terms, this means that instead of needing hundreds or thousands of coordinates to indicate how its atoms are positioned, we can achieve a fairly good description by only displaying a handful of principal components. This reduction greatly facilitates all subsequent analyses, whether qualitative or quantitative. Accordingly, PCA is used both to generate animations of the functional movements of biomolecules and to calculate their free energy surfaces or conformational entropies. However, we must note that to apply PCA effectively, it is necessary to understand its theoretical foundations in order to design strategies to avoid the algorithm's limitations. In this article, we will discuss the theoretical bases of PCA and present some procedures that allow making the most of its advantages.