The Informatics Research On Naturally Processed Peptide

Short peptides originated from endogenous proteins degraded by proteasomes are constantly presented by major histocompatibility complex (MHC) class I molecules to CD8+T cells which are ready to provide continual immune surveillance for immune system. These peptides are called naturally processed peptides or NPPs in short, which play an important role in immune tolerance, immune regulation, cell-mediated immunity, transplant rejection, personalized vaccine, and etc. Therefore, NPPs have been studied intensively, especially in the view of immunology. Naturally processed peptides can be seen as the snapshots of the translated gene products, some scholars even took them as nature’s "gene-chips". More and more researchers begin to focus on the bioinformatics research on naturally processed peptides. However, there are few special tools that can be used to predict naturally processed peptides or to explore the source of naturally processed peptides. Therefore, developing the relevant tools and software is badly in need. In this thesis, we conducted the following studies:First, design a work flow to map naturally processed peptides back to protein and nucleic sequences. This work flow is based on the BLAST method and can be used to locate naturally processed peptides to proteins and genes. After analyzing the naturally processed peptides we collected using the workflow, we confirmed the existences and expressions of quite a lot conceptual genes. Besides, we also found several new genes and novel single nucleotide polymorphism candidates.Second, develop a tool for the prediction of naturally processed peptides based on support vector machine. The tool is called NIEluter, which can be used to predict if peptides with lengths from 8 to 11 might be eluted from six HLA alleles. As many MIIC binders cannot be processed and presented in vivo, the tool for the prediction of naturally processed peptides is more helpful than those conventional predictors for MHC binders. Thus, the NIEluter can be used to predict cytotoxic T cell epitopes that can be processed and presented naturally. And then the results can be verified with experiment. This way, a lot of money and time can be saved.The workflow can be accessed via http://immunet.cn/nie/demo.html and the NIEluter is freely available at http://immunet.cn/nie/cgi-bin/nieluter.pl.