Application of Quantum Chemical Calculation for Prediction of Ultraviolet-vis Spectrum of Plant Self-protective MetabolitesProduced by UV-B Irradiation

Plants can produce various types of compounds classified as primary and secondary metabolites to survive and adapt against their given environments. These plant secondary metabolites have large chemical diversities in terms of their physicochemical properties. However, in most of the cases, information on the physicochemical properties of these compounds can be obtained by referring to articles. As it takes a tremendous amount of time due to curating published papers manually, the development of novel computational methods for prediction of these properties to shorten time with high accuracy, is needed. One of the key scientific fields, quantum chemical calculation has a potential ability because compound structures can be directly used as parameters to represent compounds' physicochemical characteristics. Thus, we aimed to develop a novel method for improving the accuracy of spectroscopic data of secondary metabolites predicted by quantum chemical calculation when comparing publicly-available data. We chose the six representative metabolites that are accumulated by UV-B irradiation for this study. The UV-vis absorption spectrum of each compound was obtained by CNDO/S semi-empirical molecular orbital calculation at the optimized structure by PM3 one. Subsequently, obtained data for several excited states were corrected by Gaussian fitting and the regression analysis was employed for these data. The absorption maximum of the corrected UV-vis spectrum was corresponded reasonably well with the reported experimental data. Our method enables us not only to shorten the data acquisition time, but also to predict spectroscopic data of every compound from the corresponding chemical structures.