Simultaneous estimation of saponin contents from soybean seed using Fourier transform infrared spectroscopy and high-performance liquid chromatography analysis

The aim of this study is to determine the feasibility of Fourier transform infrared (FT-IR) spectroscopy for simultaneous determination of saponin contents in different soybean cultivars. In cross validation between predicted content of saponin by PLS regression modeling from FT-IR spectra and measured content by HPLC, total saponin contents were predicted with good accuracy ( R 2 ≥ 0.71). In external validation, saponin group Ab ( R 2 = 0.88), saponin DDMP-group βg ( R 2 = 0.85), total saponin group B ( R 2 = 0.88), and total saponin content ( R 2 = 0.87) were predicted with good accuracy, while prediction for saponin group Aa ( R 2 = 0.58), saponin group Bb′ ( R 2 = 0.58), and total saponin group A ( R 2 = 0.25) had relatively lower accuracy. Considering these results, we suggest that the PLS prediction system for saponin contents using FT-IR analysis could be applied as a novel screening tool for high yielding lines in soybean breeding.