Abstract: Objective　In order to provide a reference basis for exploring convenient and stable fatigue markers，we studied various high-molecular-weight peptide fragments under the fatigue state and non-fatigue state using in the saliva using time of flight mass spectrometry.Methods　The saliva samples were collected from 10 healthy volunteers and the flight time of mass spectrometry was conducted using two kinds of sample treatment methods，the magnetic beads enrichment（MB） and direct detection of stock solution.This was followed by modeling via the mass spectrums of MB and supernatant（stock solution） directly collected after centrifugation.Results　Both MB and direct sampling produced good spectrograms between 2 000 and 150 00 Da，while some peaks were lost in the enrichment.The early and late spectrograms were different in each individual.Due to the limited sample size，the cross validation was not conducted，however two spectrograms were taken for each sample，and 20 early and 20 late spectrograms were used for modeling analysis.Three differential peptides was found in the two groups of stock solution samples.The cross validity of MB model was 92.06%，while that of stock solution model was 95.49%.Conclusion　There were differential peaks within the molecular weight of 2 000-15 000 Da，in different degrees of fatigue which provided a scientific basis for further realization of the convenient fatigue detection method based on biosensor technique，with important theoretical and practical significance.

Key words: Fatigue, Magnetic bead, Biomarkers, Saliva, Peptide mapping

摘要： 目的　利用目前比较成熟的唾液飞行时间质谱分析技术探索疲劳状态与非疲劳状态下有差异的大分子量肽段，为研究疲劳更加便捷稳定的标志物提供重要的理论基础。方法　采集10名健康志愿者唾液样本，采用磁珠富集（MB）和原液直接检测两种样品处理方式进行飞行时间质谱分析，以MB的唾液质谱和离心后直接取上清液（原液）的唾液质谱建模。结果　磁珠和直接采样均在2 000~15 000 Da具有很好的谱图，富集会丢失部分峰。每个人早、晚的谱图均有差异，由于样本份数有限，暂未行交叉验证，因此每个样本采两张谱图，早晚各用20张谱图进行建模分析：在原液样本两组间发现3个差异肽。MB模型的交叉效度为92.06%，原液模型的交叉效度为95.49%。结论　人不同程度疲劳状态唾液蛋白标志物的肽谱在分子量2 000~15 000 Da范围内有差异峰，为进一步实现基于生物传感器技术等的便捷疲劳检测方法提供科学依据，具有重要的理论和现实意义。

关键词: 疲劳, 磁珠, 生物标志物, 唾液, 肽图谱