Abstract
Karl Pearson developed Principal Component Analysis (PCA) in 1901 as a
mathematical equivalent of the principal axis theorem. Later on, it was given different
names according to its application in various fields. Principal Component Analysis
provides a foundation for comprehending the fundamental workings of the system
under examination. It has various applications in different fields such as signal
processing, multivariate quality control, psychology, biology, meteorological science,
noise and vibration analysis (spectral decomposition), and structural dynamics. In this
chapter, we will discuss its application in pharmaceutical research and drug discovery.
This technique allows for the representation of multidimensional data and the
evaluation of large datasets to improve data interpretation while retaining the maximum
amount of information possible. PCA is a technique that does not require extensive
computations and offers reduced memory and storage requirements. PCA can be
conceptualized as an n-dimensional ellipsoid fitted to the data, with each axis
representing a principal component. The ellipse's axes are determined by subtracting
the mean of each variable from the datasheet. In the pharmaceutical research field,
original variables are often expressed in various measurement units. Therefore, the
original variables are divided by their standard deviation once the mean has been
subtracted. This step is taken to work with z-scores, which are further used for
extracting the eigenvalues and eigenvectors of the original data.