Chapter 13 Principal Components Analysis: A Review of its Application on Molecular Dynamics Data

Publisher Summary This chapter discusses the need for multivariate analysis in biophysical studies, presents the way principal components analysis (PCA) can be implemented in the analysis of molecular dynamics data, and provides insights into the pitfalls and common errors associated with multivariate techniques. Principal components analysis (PCA) is the simplest of multivariate techniques that is used to reduce or simplify large and complicated sets of data. PCA is predicated on the assumption that the phenomena of interest can be explained by the variances and covariances between the p variables in the original data set. PCA is used in a wide array of applications ranging from crystallographic and nuclear magnetic resonance (NMR) structure ensembles, protein and peptide folding/unfolding, the structural determinants of transmembrane proteins and channels, large-scale domain motion, and many more. PCA has been shown to be a powerful tool in evaluation of DNA flexibility in molecular dynamics simulations. The chapter highlights the utility of PCA to reduce the complexity of variables describing the dynamics data. PCA and the mathematically identical essential dynamics have proved useful in the detection of important motions in biomolecules ranging from proteins to nucleic acids.