Principal Component Analysis (PCA) is a widely used exploratory and visualization technique. PCA tries to describe the important variability in the MS data in a reduced number of dimensions, i.e. original features are transformed into a smaller number of important directions called principal components (PCs). The first principal component accounts for as much of data variability as possible, the second component (orthogonal to the first one) accounts for as much of the remaining variability as possible, etc.

In the Tutorial we show how to perform PCA (Principal Component Analysis) in the Spectrolyzer software. It is assumed that raw MS data have already been prepared for statistical analysis, i.e. all necessary preprocessing steps have been completed, including Peak Extraction (see Step 2) and Binning (see Step 3).

A number of interactive 2D and 3D visualization tools is used to facilitate interpretation of PCA results. In particular, scree plots as well as available score and loading scatter plots can be used to infer a number of important dimensions in the data and to assess importance of consecutive features (bins). Using PCA we can also investigate ability of our data to separate group (i.e. ability to discriminate between Cancer and Control samples).

Note that  PCA analysis is only one (simple) example of statistical methods available in the Spectrolyzer. Please refer to our Manual if you want to know more about methods available in the Statistical Analysis module, including advanced approaches to biomarker discovery, classification or dimension reduction.