UHPLC‐Q‐TOF‐MS Phytochemical Characterization and ATR-FTIR analysis Combined with Machine Learning for Asparagus racemosus Leaves

The medicinal plant Asparagus racemosus Wild., has been widely used in traditional medicine to cure various diseases such as cough, diabetes, gastric issues, diarrhoea, headaches, and rheumatism. This paper studies phytochemicals of the species native to Gia Lai province, Vietnam, before its conservation and cultivation. A. racemosus leaves were extracted with water then UHPLC-Q-TOF-MS spectroscopy was used to screen for their phytochemicals. In water extract of A. racemosus leaves, UHPLC-Q-TOF-MS tentatively identified two flavonoids (Quercetin-3-glucuronide, Rutin), five steroidal saponins (Shatavarin I, Shatavarin IV, Shatavarin IX, Asparacoside, Asparanin A), and two steroids (β-sitosterol, Daucosterol). One objective of this study was to identify and classify the chemical differences between leaves of Asparagus racemosus collected in different places of Gia Lai using Attenuated Total Reflection - Fourier Transform Infrared spectroscopy (ATR-FTIR) combined with machine learning algorithms. A total of 40 samples of A. racemosus’s leaf powder were analyzed using ATR-FTIR spectroscopy. The Savitzky-Golay algorithm, unsupervised (Hierarchical Clustering) and supervised (Principal Component Analysis and Support Vector Machine) machine learning algorithms were employed to classify the ATR-FTIR spectra. The combination of ATR-FTIR spectral data and machine learning algorithms successfully discriminated and classified the collected places of A. racemosus leaves. The experimental findings confirm the A. racemosus species for conservation and cultivation in Vietnam and contribute the benefits to the chemical literature of Vietnamese natural flora. A. racemosus should be further studied for its pharmaceutical activities.