5 – Protein inference in shotgun proteomics

Proteomics is a rapidly developing research area whose main objective is to conduct a large-scale study of proteins expressed in an organism. Currently, shotgun proteomics is probably the most widely used method in practice, which identifies complex protein mixtures using a combination of high-performance liquid chromatography and mass spectrometry. There are two key computational problems in the protein identification process: peptide identification and protein inference. In this chapter, we first present the protein inference problem and then illustrate how to solve this problem using different data mining techniques. Briefly, the protein inference problem can be modeled as a regression problem, a classification problem, or a cluster analysis problem. The multiple views of the same problem may provide deep insights on how to use data mining methods to solve real bioinformatics problems. We additionally present two different validation methods for this issue. Finally, we provide some open problems in this field, which we leave to the readers.