在代谢组学研究方法中，对一组具有生物学意义且平行的核磁共振波谱数据进行处理是其方法研究的关键，其中包含谱峰对齐、谱峰提取、噪音去除等几个步骤。该文提出了一种全新且快速有效的核磁共振波谱批量处理方法。其原理主要为：在保持波谱形状不发生变化的同时，通过平移目标谱使其与参考谱达到最大程度的相关，从而实现谱峰对齐及利用谱峰最大值的判断来实现谱峰变量的提取。此外，该文还结合大鼠尿样和血样的分析结果，详细地说明了算法的准确性及其它优点。该文所涉及的算法是基于Matlab程序进行编译的，可以结合不同实验进行改编，具有良好的扩展性。%The method for NMR spectra post processing, especially for a set of parallel biological NMR spectra, is crucial in metabolomics studies. Here, an efficient spectra post processing method for peak alignment and peak extraction is proposed, which not only did well in accurate results with high resolution but also had advantages in batch processing and time consumption. The spectrum alignment was completed with the shift of the spectra without changing the profiles of the NMR spectra, and the results were evaluated by the regression coefficientR. The peak extraction step was achieved by repeated searching the maximum value in the NMR spectrum, and was a prerequisite for metabolomics analysis. The extraction of all relevant peaks contained in the complex mixture spectra, rid of any non-significant signal could be easily applied to the statistical analysis. This new approach was applied to a set of1H NMR spectra of rat plasma and urine to demonstrate the efficiency of the method. The whole theory is compiled in Matlab, and the implementation code is available upon request.
展开▼
