Objective To establish a rapid prediction method of the antioxidant activity in aqueous extract solutions of Melastoma dodecandrum based on ultraviolet spectroscopy and partial least squares regression algorithm. Methods The DPPH free radical scavenging effect was used to characterize the antioxidant activity of aqueous extract solutions of Melastoma dodecandrum. The ultraviolet spectra of 190-600 nm were collected. The partial least squares regression model of antioxidant activity was established after optimizing the wavelength range and preprocessing method. The software was devised using Visual Basic as the integrated development environment to provide a convenient tool for the rapid determination of antioxidant activity. Results The optimal partial least squares regression model was established based on 200-290 nm as wavelength range and unit variance scaling as preprocessing method. The correlation coefficient of calibration, root mean square error of estimation, root mean square error of cross-validation was 0.887, 2.20% and 2.17%, respectively. The correlation coefficient of validation, root mean square error of prediction was 0.868, 2.08%. The average predicted recovery was 100.1±2.3%. With the predictive function in the software, the antioxidant activity of aqueous extract solution of Melastoma dodecandrum can be calculated automatically within 2 s after collecting the ultraviolet spectra. Conclusions This study provides a rapid method for the prediction of antioxidant activity in aqueous extract solutions of Melastoma dodecandrum.

OBJECTIVE： To establish a method for simultaneous determination of 4 triglyceride anti-tumor components in Coix lacryma seed oil. METHODS： HPLC-ELSD was adopted. The determination was performed on Inertsil ODS-3 C18 column with mobile phase consisted of acetonitrile-isopropanol （57 ∶ 43， V/V） at the flow rate of 1.0 mL/min. The column temperature was 30 ℃， and the sample size was 10 μL. Evaporative light scattering detector was used， the drift tube temperature was 70 ℃， and the gas flow rate was 2 L/min. Using glycerol trioleateas internal standard， relative correction factors （RCF） of linolein trilinolein， 1，2-linoleic acid-3-palmitic acid glyceride and 1-palmitic acid-2-oleic acid-3-linoleic acid glyceride were calculated respectively. The contents of above 3 components in C. lacryma seed oil were calculated by RCF. The contents of 4 components in C. lacryma seed oil were determined by external standard. The results of content determination by quatitative analysis of multi-components by single marker （QAMS） were compared with external standard method. RESULTS： The linear ranges were 0.15-4.50 μg for linolein trilinolein， 0.15-4.50 μg for 1，2-linoleic acid-3-palmitate， 0.35-10.50 μg for 1-palmitic acid-2-oleic acid-3-linoleic acid glyceride， 0.35-10.50 μg for glycerol trioleate （r≥0.999 5）. The limits of quantification were 0.13， 0.06， 0.07， 0.12 μg. The limits of detection were 0.04， 0.02， 0.02， 0.03 μg， respectively. RSDs of precision， stability， and repeatability tests were less than 2.0％（n＝6）. The average recoveries were 95.43％-102.67％（RSD＜2.0％， n＝6）. Average RCFs of linolein trilinolein， 1，2- linoleic acid-3-palmitic acid glyceride and 1-palmitic acid-2- oleic acid-3-linoleic acid glyceride were 0.31， 0.88， and 1.21， respectively. RCFs reproducibility was perfect under different experiment conditions. There was no significant difference in results of content determination between QAMS and external standard method （P＞0.05）. CONCLUSIONS： The method is simple， rapid， accurate and reliable. It is used for simultaneous determination of linolein trilinolein， 1， 2-linoleic acid-3-palmitate， 1-palmitic acid-2-oleic acid-3-linoleic acid glyceride and glycerol trioleateas in C. lacryma seed oil.