近红外技术结合参数轨迹策略快速评价土茯苓质量
A processing-trajectory strategy for rapid quality assessment of Rhizoma smilacis Glabrae by NIR combined with PLS modeling
吕蒙莹 1万夏芸 2帅锦浩 2王杨 2刘晓庆 3严欢 4赵娜3
作者信息
- 1. 扬州大学医学院(转化医学研究院),江苏扬州 225009;扬州大学附属医院/国家中医药管理局胃癌辨治重点实验室,江苏扬州 225009;扬州大学广陵学院,江苏扬州 225000
- 2. 扬州大学医学院(转化医学研究院),江苏扬州 225009;扬州大学附属医院/国家中医药管理局胃癌辨治重点实验室,江苏扬州 225009
- 3. 石河子大学药学院/新疆植物药物资源与利用教育部重点实验室,新疆石河子 832002
- 4. 新疆维吾尔自治区分析测试研究院,乌鲁木齐 830063
- 折叠
摘要
为建立一种能快速测定土茯苓中落新妇苷含量的方法,运用近红外漫反射光谱技术(NIRs)采集不同产地土茯苓NIR光谱,并采用高效液相色谱(HPLC)法测定土茯苓中主要生物活性成分落新妇苷含量.利用偏最小二乘法(PLS),结合NIR光谱和有效成分含量建立落新妇苷的定量模型.针对落新妇苷PLS模型的参数优化,提出一种参数轨迹优化策略,该策略集成4种预处理方法、潜变量因子数设置为1～10、间隔偏最小二乘法(iPLS)和反向间隔偏最小二乘法(BiPLS)的变量筛选方法.结果表明:参数轨迹策略比常用的分步策略更高效准确,通过新策略得到3个预测性能较好的模型.此外,还建立了偏最小二乘判别分析(PLS-DA)模型来区分土茯苓样品的地理来源.综上,NIR结合PLS模型的参数轨迹策略可作为一种快速、可靠和经济的原料检测和产品质量控制方法.
Abstract
To establish a method for rapid determination of astilbin in Smilax glabra by near infrared spectroscopy(NIRS)combined with processing-trajectory strategy,so as to evaluate the quality of Smilax glabra from different pro-duction areas.NIR spectra of Smilax glabra from different production areas were collected by near infrared spectroscopy,and the content of astilbin,and the main bioactive component in Smilax glabra,was determined by high performance liq-uid chromatography(HPLC).The quantitative model of astilbin was established by partial least squares(PLS)combined with the collected NIR spectra and the content of astilbin.For the parameter optimization of astilbin PLS model,a pro-cessing-trajectory strategy that integrated four pretreatment methods,latent factors setting from 1 to 10 and interval par-tial least squares regression(iPLS),backward interval partial least squares(BiPLS)-based variable selection was pro-posed.The results showed that the processing-trajectory strategy proved to be more efficient and accurate than the com-monly used step-by-step strategy.Three good models with better predictive performance were obtained through the new strategy.Furthermore,PLS-DA model was built to differentiate the geographical origin of the Rhizoma smilacis Glabrae samples.NIR coupled with the novel processing-trajectory strategy of PLS model can be a rapid,reliable and cost-effec-tive method for raw material testing and product quality control.
关键词
土茯苓/近红外光谱/偏最小二乘法/参数轨迹策略/质量控制Key words
Rhizoma smilacis Glabrae/near infrared spectroscopy/partial least squares/processing-trajectory strategy引用本文复制引用
基金项目
国家自然科学基金资助项目(81903906)
国家自然科学基金资助项目(81960769)
江苏省高等学校自然科学基金面上项目(23KJB360018)
中国博士后科学基金资助项目(2018M642346)
江苏省中医药管理局科研项目(MS2022094)
国家留学基金委员会资助项目(201809300008)
扬州大学"青蓝工程"优秀骨干青年教师项目(2023-08)
扬州大学人才引进基金项目(2021-08)
扬州大学技术创新与培育基金项目(2019CXJ175)
出版年
2024
国家科技期刊平台
NSTL
万方数据