目的 探究草果Tsaoko Fructus仁不同炮制品（清炒草果仁、姜制草果仁和麸炒草果仁）的气味特征和挥发性成分之间的差异与特点。方法 采用电子鼻联合顶空-气相色谱-质谱法（headspace-gas chromatography-mass spectrometry，HS-GC-MS）对清炒草果仁、姜制草果仁和麸炒草果仁共计30批草果仁不同炮制品中的成分进行检测，利用主成分分析（principal component analysis，PCA）、Loadings和绘制气味雷达图等方法分析电子鼻信号响应结果，同时对HS-GC-MS鉴定出的化合物做正交偏最小二乘法-判别分析（orthogonal partial least squares-discriminant analysis，OPLS-DA），以筛选变量重要性投影（variable importance projection，VIP）值＞1为标准筛选潜在差异性成分。结果 电子鼻数据分析结果显示，气味差异主要集中在传感器W5S、W1W、W2W，表明草果仁不同炮制品在气味特征上存在差异。HS-GC-MS鉴定出66个化合物，其中共有化合物35个，进一步通过OPLS-DA对清炒草果仁、姜制草果仁和麸炒草果仁进行两两比较，筛选出差异成分25个。根据对挥发性差异成分与气味进行的Pearson相关性分析，结果显示区分草果仁3种不同炮制品的基础物质主要包括桉叶油醇、邻伞花烃等。据此推测，部分醇类、萜烯类、醛类和酚类成分可能通过引起传感器W5S、W1W、W2W的变化，进而实现对草果仁不同炮制品的区分。相关性热图分析结果进一步证实了这一推测。经查阅文献，选取了水芹烯、桉叶油醇、芳樟醇、萜品油烯、α-松油醇、柠檬醛、2,3-二氢-1H-茚-4-甲醛、甲基庚烯酮和月桂烯9种成分，通过系统的研究和对比分析，揭示了这9种成分与草果仁不同炮制品功效的关联。结论 采用多元统计分析方法，研究草果仁3种不同炮制品-传感器响应值-挥发性成分之间的相关性，探究草果仁不同炮制品挥发性成分的差异，为草果仁不同炮制品临床应用提供新思路。

Objective To explore the differences and characteristics between the odor characteristics and volatile components of different processed products of seeds of Caoguo (Tsaoko Fructus) [stir-fried seeds of Tsaoko Fructus (qcTF), ginger made seeds of Tsaoko Fructus (jzTF), stir fried seeds of Tsaoko Fructus with bran (fcTF)]. Methods The electronic nose combined with headspace-gas chromatography-mass spectrometry (HS-GC-MS) was used to detect the components in 30 batches of different processed products from qcTF, jzTF and fcTF. Principal component analysis (PCA), loadings, and odor radar mapping were used to analyze the electronic nose signal response results, simultaneously, performing orthogonal partial least squares-discriminant analysis (OPLS-DA) on the compounds identified by HS-GC-MS to screen for potential differential components based on variable importance projection (VIP) values > 1. Results The analysis results of electronic nose data showed that the odor differences were mainly concentrated in sensors W5S/W1W/W2W, indicating that there were differences in the odor characteristics of seeds of Tsaoko Fructus from different processed products. 66 Compounds were identified by HS-GC-MS, among them there were 35 common compounds. Further, OPLS-DA was used to compare the qcTF, jzTF, and fcTF, and 25 differential components were screened. According to Pearson correlation analysis of volatile differential components and odors, the results showed that the basic substances used to distinguish the three different processed products mainly included cineole, o-cymene, etc. Based on this, it was speculated that some alcohols, terpenes, aldehydes, and phenols may differentiate different processed products of seeds of Tsaoko Fructus by causing changes in sensors W5S/W1W/W2W. The subsequent correlation heatmap analysis further confirmed this speculation. After consulting literature, a total of nine components including phellandrene, cineole, linalool, terpinene, α-terpineol, citral, 2,3-dihydro-1H-indene-4-formaldehyde, methyl heptenone and myrcene were selected. Through systematic research and comparative analysis, the correlation between these nine components and the efficacy of different processed products of seeds of Tsaoko Fructus was revealed. Conclusion Multiple statistical analysis was used to study the correlation between the sensor response values and volatile components of three different processed products of seeds of Tsaoko Fructus, and to explore the differences in volatile components of different processed products of seeds of Tsaoko Fructus. This provides new ideas for the clinical application of different processed products of seeds of Tsaoko Fructus.

R283.6

河北省省级科技计划项目（23372502D）；河北省自然科学基金资助项目（H2022423335）；河北省中医药管理局科研计划项目（2024104）