Fig. 4

Principal component analysis (PCA) plot of individuals and variables. Principal components (PCs) 1 (representing 38.6% of explained variance) and PC2 (15.5% of explained variance) are plotted. (a) PCA biplot of individual samples in relation to PC1 and PC2. A high cos2 indicates a quality of good representation of the individual on the PC. (b) Correlation circle of a PCA. In a PCA correlation circle plot, vector direction and length play crucial roles in conveying information about the relationships between variables and principal components. The direction of a vector signifies the strength and nature of the correlation between an original variable and a specific principal component; proximity to the component indicates a stronger correlation. Meanwhile, vector length is equally significant, as longer vectors indicate higher correlations. This means that variables with longer vectors contribute more significantly to the variability explained by that principal component. Interpreting the angle between vectors provides insights into the relationships between variables: smaller angles denote positive correlations, while larger angles imply weaker or non-existent correlations.