An improved scoring method for the identification of endogenous peptides based on the Mascot MS/MS ion search
An improved scoring method for the identification of endogenous peptides based on the Mascot MS/MS ion search
复制标题
DOI：
10.1039/c8an02141d
复制
发表时间：
2019-05-07
期刊：
影响因子：
4.2
通讯作者：
Chen, Yet-Ran
中科院分区：
文献类型：
--
作者：
Chen, Ying-Lan;Chang, Wei-Hung;Chen, Yet-Ran
为了利用MS/MS分析和对多肽序列数据库进行搜索来鉴定内源性肽，需要考虑所有可能的蛋白水解裂解的非酶特异性（NES）搜索。然而，使用NES搜索比特定酶搜索产生更多的假阳性命中，因此显示较低的识别性能。在本研究中，研究了使用子排序匹配来提高Mascot NES搜索的识别性能，并开发了一种新的评分方法，该方法考虑了所有子排序随机匹配概率的贡献，称为贡献分数（CS）。与使用Mascot第一评分和delta评分（DS）相比，使用完整蛋白质数据库的Mascot NES搜索的CS显示出最高的识别灵敏度。通过DS和CS鉴定出的自信肽是互补的。应用于植物内肽鉴定时，DS和CS对番茄内肽的识别率分别为176.3%和184.2%，高于Mascot排名第一的识别率。与使用Mascot和DS相比，DS和CS组合鉴定的番茄内源肽分别增加200.0%和8.6%。这种结合CS和DS的方法可以显著提高内源性肽的鉴定性能，无需复杂的计算步骤，也可以提高酶特异性搜索的鉴定性能。该方法除可用于植物肽组学分析外，还可用于改进不同物种的肽组学研究。本文开发了一个基于Mascot搜索结果计算DS和CS的web界面。
To identify endogenous peptides using MS/MS analysis and searching against a polypeptide sequence database, a non-enzyme specific (NES) search considering all of the possible proteolytic cleavages is required. However, the use of a NES search generates more false positive hits than an enzyme specific search, and therefore shows lower identification performance. In this study, the use of the sub-ranked matches for improving the identification performance of the Mascot NES search was investigated and a new scoring method was developed that considered the contribution of all sub-ranked random match probabilities, named the contribution score (CS). The CS showed the highest identification sensitivity using the Mascot NES search with a full protein database when compared to the use of the Mascot first ranked score and the delta score (DS). The confident peptides identified by DS and CS were shown to be complementary. When applied to plant endogenous peptide identification, the identification numbers of tomato endogenous peptides using DS and CS were 176.3% and 184.2%, respectively, higher than the use of the first ranked score of Mascot. The combination of DS and CS identified 200.0% and 8.6% more tomato endogenous peptides compared to the use of Mascot and DS, respectively. This method by combining the CS and DS can significantly improve the identification performance of endogenous peptides without complex computational steps and is also able to improve the identification performance of the enzyme specific search. In addition to the application in the plant peptidomics analysis, this method may be applied to the improvement of peptidomics studies in different species. A web interface for calculating the DS and CS based on Mascot search results was developed herein.