Wieczorek Samuel, Combes Florence, Borges Hélène, Burger Thomas
Université Grenoble Alpes, CEA, Inserm, BGE U1038, Grenoble, France.
CNRS, BIG-BGE, Grenoble, France.
Methods Mol Biol. 2019;1959:225-246. doi: 10.1007/978-1-4939-9164-8_15.
ProStaR is a software tool dedicated to differential analysis in label-free quantitative proteomics. Practically, once biological samples have been analyzed by bottom-up mass spectrometry-based proteomics, the raw mass spectrometer outputs are processed by bioinformatics tools, so as to identify peptides and quantify them, by means of precursor ion chromatogram integration. Then, it is classical to use these peptide-level pieces of information to derive the identity and quantity of the sample proteins before proceeding with refined statistical processing at protein-level, so as to bring out proteins which abundance is significantly different between different groups of samples. To achieve this statistical step, it is possible to rely on ProStaR, which allows the user to (1) load correctly formatted data, (2) clean them by means of various filters, (3) normalize the sample batches, (4) impute the missing values, (5) perform null hypothesis significance testing, (6) check the well-calibration of the resulting p-values, (7) select a subset of differentially abundant proteins according to some false discovery rate, and (8) contextualize these selected proteins into the Gene Ontology. This chapter provides a detailed protocol on how to perform these eight processing steps with ProStaR.
ProStaR是一款专门用于无标记定量蛋白质组学差异分析的软件工具。实际上，一旦通过基于自下而上质谱的蛋白质组学对生物样品进行了分析，质谱仪的原始输出就会由生物信息学工具进行处理，以便通过前体离子色谱图积分来鉴定肽段并对其进行定量。然后，在进行蛋白质水平的精细统计处理之前，使用这些肽段水平的信息来推导样品蛋白质的身份和数量是很常见的做法，以便找出在不同样品组之间丰度有显著差异的蛋白质。为了完成这一统计步骤，可以依靠ProStaR，它允许用户：(1)加载格式正确的数据；(2)通过各种过滤器对数据进行清理；(3)对样品批次进行归一化；(4)插补缺失值；(5)进行零假设显著性检验；(6)检查所得p值的校准情况；(7)根据一定的错误发现率选择差异丰富蛋白质的子集；(8)将这些选定的蛋白质纳入基因本体论中进行背景分析。本章提供了一份关于如何使用ProStaR执行这八个处理步骤的详细方案。
