3.3. FT-IR spectrum data and multivariate statistical analysis in Citrus
To better understand FT-IR spectra variability, PCA of FT-IR was
performed by generating two-dimensional plots between controland mutant Citrus samples using two principal components, PC1
and PC2. The PCA analysis result from FT-IR spectra data showed
that the overall variation scores of PC1 and PC2 were 63.4% and
15.5%, respectively. This result indicates that overall variability
was 78.9% when PC1 and PC2 were combined (Fig. 2A). The PCA
score plot with a high explanation power of PC1 and PC2 was
examined, and the scatter plots comparing C. unshiu Marc. Cv.
Miyagawa-Wase and the mutagenized M lines were divided into
two parts, with the top being high and the bottom being low, by
the PC2 score Based on the PCA plot, C. unshiu Marc. Cv.
Miyagawa-Wase lines (control) were broadly located in PC1 but
located in the middle of the PC2 axis, and mutagenized M2 lines
that had similar sugar contents as the control were closely located
together in the C. unshiu Marc. Cv. Miyagawa-Wase group. M3 and
M4 lines were individually located in the bottom and top parts
based on their PC2 scores. M1 lines, containing the highest sugar
contents were clustered between M2 and M3 lines (Fig. 2A). To
understand biological information based on the spatial distribution
of position from the PCA score plot in Citrus sample lines, we calculated correlations based on the sugar and acid content difference
(Table 1) and sample locations in the PCA score plot. The plot
revealed that C. unshiu Marc. Cv. Miyagawa-Wase with its relatively low sugar content was separately located in the middle of
the Y-axis. However, mutagenized M1 lines with the highest sugar
content were located in a separate cluster between the M2 and M3