目的 建立多指标定量与化学计量学及Logistic回归分析的整合分析方法，评价不同产地白英Solanum lyratum的质量差异。方法 采用Purospher STAR RP₁₈色谱柱，流动相为乙腈-0.5%磷酸，梯度洗脱，检测波长为254 nm（原儿茶酸、绿原酸、香草酸、咖啡酸、阿魏酸、芦丁、大豆苷、蒙花苷和16-妊娠双烯醇酮）和210 nm（延龄草苷、去半乳糖替告皂苷和薯蓣皂苷元）。结合多指标定量分析、化学计量学及Logistic回归分析对48批白英样品进行质量差异评价。结果 含量测定的方法学验证均良好，定量分析的12个成分线性关系、分离度均良好，平均加样回收率为96.86%～99.47%，RSD均＜2.0%；化学计量学分析结果显示不同产地白英的区分度较大，48批样品分为3类，延龄草苷、芦丁、16-妊娠双烯醇酮、绿原酸和薯蓣皂苷元的变量权重值（variable importance projection，VIP）值＞1；因子分析法结果显示48批白英的综合得分为−1.351～1.190，其中S42综合得分最高。Logistic回归分析结果与因子分析的分级结果一致。结论 基于多指标定量分析方法操作简便，结果准确，结合化学计量学及Logistic回归分析手段可实现对不同产地白英的质量分析与评价，可以其质量控制与评价提供参考。

Objective To establish an integrated analysis method of multi-index quantitative analysis, chemometrics and Logistic regression analysis, and to evaluate the quality differences of Baiying (Solani Lyrati Herba) from different production areas. Methods Analysis was performed on Purospher STAR RP₁₈ column with the mobile phase of acetonitrile-0.5% phosphoric acid solution for gradient elution. The detection wavelength were set at 254 nm for protocatechuic acid, chlorogenic acid, vanillic acid, caffeic acid, ferulic acid, rutoside, daidzin, buddleoside, 16-dehydropregnenolone, 210 nm for trillin, uttronin A and diosgenin. The quality difference of 48 batches of Solani Lyrati Herba samples was evaluated by multi-index quantitative analysis, chemometrics and Logistic regression analysis. Results The determination methods were well verified, the resolution and linear relationship of 12 components in quantitative analysis were good, the average recoveries were 96.86%—99.47% with RSDs＜2.0%. The results of chemometrics analysis showed that the discrimination of Solani Lyrati Herba from different producing areas was large, and 48 batches of samples were divided into three categories. The VIP values of trillin, rutoside, 16-dehydropregnenolone, chlorogenic acid and diosgenin were > 1. The results of factor analysis showed that the comprehensive scores of 48 batches of Solani Lyrati Herba were between -1.351 and 1.190, and the comprehensive score of S42 was the highest. The results of Logistic regression analysis were consistent with the classification results of factor analysis. Conclusion Based on the multi-index quantitative analysis method, the operation is simple and the result is accurate. Combined with chemometrics and Logistic regression analysis, the quality analysis and evaluation of Solani Lyrati Herba from different producing areas can be realized, which can provide reference for its quality control and evaluation.

R286.2

国家中医药管理局全国中药特色技术传承人才培训项目［国中医药人教函（2023）96号］