Scatter plot of sample-based PCA loadings. Visualization of PCA loadings for all marker candidates of the experiment. Loadings were calculated according to the first two principal components of sample-based PCA. Black, green and blue dots represent unidentified marker candidates. Green and blue dots correspond to candidates of clusters 14–15 and 27–33, respectively. Red asterisks represent identified markers. Marker abbreviations are explained in section "Application of 1D-SOM" and in table 2.