The Prediction of Peptide Detectability in MS Data Analysis Using Logistic Regression

The probability of the peptide that can be observed in the proteomics experiment based on mass spectrometry (MS) is not only determined by the abundance of proteins, but also heavily determined by the properties or structures of peptides. The set of peptides that are detected from a single protein could differ from one experiment to another substantially. We present an approach to predict the probability of the peptide that can be detected in MS-based proteomic experiment based on the logistic regression using the properties of peptides, and it has been tested and verified on the different datasets and showed satisfactory performance.