云南地方鸡肌肉化学成分近红外特征光谱产生机理及校正模型研究

Near infrared spectroscopy analytical technology is a new detection methodology for chicken. However, there are some problems as follows: the mechanism of near infrared characteristic spectra based on chemical compositions of chicken muscle is not clear; accuracy and stability of calibration models on chicken are insufficient effectively; nobody has researched on near infrared spectra of Yunnan native chicken. Taking leg muscles and breast muscles of six representative Yunnan native chickens as research objects, this project is to research the correlation of near infrared spectra (NIRS) between pure chemical reagents (amino acid and fatty acid) and their molecule groups, the similarities of NIRS between pure chemical reagents and muscle extracts( protein extracts and fat extracts) of chicken, the similarities and differences between NIRS of muscle extracts and freeze-dried powders from chicken, the similarities and differences of NIRS between leg muscles and breast muscles from muscle extracts and freeze-dried powder of chicken，the correlation between content and NIRS from chemical compositions of chicken muscle. Through the association analysis of chemical compositions, muscle extracts and freeze-dried powders of chicken, near infrared characteristic spectra of chicken will be selected. The chemical essence and formation mechanism of near infrared characteristic spectra of chicken muscle will be explained. The characteristic calibration models of chicken chemical compositions will be built based on characteristic spectra selection, spectral preprocessing and sample sets optimization. The results will elucidate the formation mechanism of near infrared characteristic spectra and build calibration models based on chemical compositions of muscle in Yunnan native chicken. It will provide a new ideas and solutions to improve the accuracy and stability of near infrared analysis, which have important academic value and broad application prospects.

近红外分析是鸡肉检测新方法，目前存在鸡肉化学成分特征光谱产生机理不明确，校正模型准确性和稳定性不足，云南地方鸡近红外研究尚属空白等问题。以6种有代表性的云南地方鸡腿肌、胸肌为研究对象，研究鸡肉中化学成分(氨基酸、脂肪酸)纯品与其分子基团NIRS的相关性，化学成分纯品与鸡肉蛋白质、脂肪提取物NIRS谱带的相似性，鸡肉提取物与冻干粉NIRS谱带的异同性，两类样品中腿肌和胸肌NIRS的异同性，鸡肉化学成分含量与NIRS间的相关性，将化学成分纯品-鸡肉化学成分提取物-鸡肉三个层次相关联，获得基于鸡肉化学成分的NIRS特征光谱，探明特征吸收光谱的化学本质和形成机制；同时进行建模特征谱区筛选，光谱预处理及样本集优化，建立鸡肉化学成分特征校正模型。研究结果将阐明鸡肉化学成分NIRS特征吸收光谱的产生机理，为有效提高鸡肉近红外分析的准确性和稳定性提供一条新的解决途径，具有重要的理论价值和广泛的应用前景。

NIRS是鸡肉检测新方法，目前存在鸡肉化学成分特征光谱产生机理不明，模型准确性和稳定性不足，云南地方鸡肌肉NIRS研究尚属空白等问题。本项目首先研究了20种氨基酸和16种脂肪酸纯品NIRS与其分子基团NIRS，氨基酸纯品NIRS与云南地方鸡肌肉蛋白质提取物NIRS，脂肪酸纯品NIRS与云南地方鸡肌肉脂肪提取物NIRS，提取物NIRS与冻干粉NIRS之间的相关性，提取物和冻干粉中15种氨基酸和7种脂肪酸含量差异对鸡肉NIRS建模的影响，揭示了鸡肉化学成分分子结构及其NIR特征光谱与鸡肉冻干粉NIRS之间的内在联系，阐明了鸡肉化学成分特征吸收光谱的化学本质。蛋白质提取物NIRS与20种氨基酸纯品NIRS无完全重合，6个明显吸收峰均与不同氨基酸纯品中的某些特征吸收峰重合。脂肪提取物NIRS与脂肪酸纯品NIRS在1000-1900nm基本一致，1900-2210nm谱区上多出三个小的吸收峰。冻干粉NIRS与蛋白质提取物NIRS谱形相近，与脂肪提取物NIRS差异较大。冻干粉在1186-1192nm、1512-1522nm、1729-1740nm、2057-2062nm、2169-2180nm、2294-2306nm六个谱区上的吸收峰与蛋白质提取物的一致，冻干粉中1186-1192nm、1729-1740nm、2294-2306nm三个谱区上的光谱还受到脂肪吸收峰的影响。其次，以5种云南地方鸡腿、胸肌冻干粉中粗蛋白、粗脂肪、15种氨基酸、7种脂肪酸等为研究对象，进行了建模特征光谱，光谱预处理方法、异常样本筛选、建模校正集的优化研究，建立了上述24种化学成分基于鸡肉腿肌和胸肌冻干粉的优化特征校正模型，并以另外3种云南地方鸡冻干粉进行验证，研究了鸡种对模型的影响。在特征光谱范围内使用原始光谱所建模型更优，腿胸肌模型的适用性更广，粗蛋白和氨基酸的建模特征光谱位于1000-1900nm，粗脂肪和脂肪酸的特征光谱位于1000-1850nm。粗蛋白、粗脂肪、赖氨酸、丙氨酸、缬氨酸、异亮氨酸、棕榈酸C16:0、油酸C18:1n-9c模型达到实用。二次外部验证可知，瓢鸡和茶花鸡可以使用粗蛋白和粗脂肪模型进行预测，茶花鸡2号的预测效果较差。15种氨基酸和7种脂肪酸的预测模型这三种鸡均不适用。可见，鸡种不同，模型的适用性差异较大。