New Protein Discovery Algorithm Based On The Experimental Data Of Mass-Spectrometry

It is a commonly used method to identify protein combine tandem mass spectrometry with database searching in proteomics. The database searching strategy increases the analytical efficiency of biological mass spectrometry data. But it's hard to completely solve all the problems of protein identification because the complexity of the experimental process of MS, the diversity of biological protein samples, and the limitations of the database searching algorithm. With the attention of new proteins, how to improve the accuracy of the identification of new protein is the most important problem that must be solved. To explore the problem of new protein fully, common strategies of protein identificaton and the existing problems, the database searching strategy, quality control method for peptide, ambiguous matches and group set model about it are considered. In a word, the work start with the following two aspects.Firstly, find ways to improve the accuracy of identification of new proteins from the two level of peptide and protein. At the peptide level, through the evaluation of the quality control method of the multi-parameter integration, the experiment was designed to observe and compare the results of different parameters, and a new quality control method was put forward. Two standard protein data sets were selected to evaluate the accuracy of new quelity control method, and a new procedure was successfully constructed and more new peptide was found in the two standard protein data. At the protein level, group model, a simple set model was proposed. Goup model retains all the positive results and classifies the high credibility of protein which reducing the influence of accuracy by ambiguous matches.Secondly, the protein novelty annotation process was designed according to the database step by step search strategy. Together with the new method for quality control of peptide, which can be used for the discovery of new protein in MS data. Four real data sets was selected and 5611 new proteins were identified and accuracy increased by 19.4%. From the final results, the new protein discovery algorithm based on the experimental data of mass-spectrometry is effective and can identify more peptides and high confidence protein from the MS data.