An Algorithm Identifying Mutations And Post-translational Modifications Of Proteins Based On MS/MS

With accomplishment of human genome project, the research of life sciences enters on post genome era. The core of life sciences during post genome era is to reveal expressing rules and biology functions of all proteins expressed by genes,which in fact carry out life movements. Therefore, the core of life sciences transfers from genomics to proteomics. Identifying sequences of proteins is a very significant task in proteomics.The laboratories and companys researching bioinformatics in the world have put forward many algorithms based on MS/MS so far,which are capable of identifying proteins. However, only a few algorithms can identify the proteins containing mutations or post-translational modifications, and these algorithms also have many limitations. Post-translational modifications play important roles in organisms, which make structures of proteins more complicated and functions more consummate. Most proteins contain post-translational modifications, and there are more than 200 types of covalent modifications on the residue of amino acids.In addition, mutations of proteins also play important roles in organisms, which are caused by mutations of genes. Therefore, identification of proteins that contain mutations or post-translational modifications is an invaluable research of life sciences, which is also full of challenges.An algorithm named Spectral Alignment was put forward in 2000 by Pavel A.Pevzner. The algorithm could identify proteins containing mutations or post-translational modifications, and did not have to know about the types of mutations or post-translational modifications MS/MS contain. However, the algorithm had some disadvantages which lead to low correctness. This paper put forward an algorithm named Modified Spectral Alignment,which has some improvement on the basis of Spectral Alignment.The algorithm not only betters the way to compute D(k)and k-optimal Spectral Alignment of original algorithm, inducts Relative Entropy Principle to improve precision of scoring function, takes into account simultaneous analysis of N-terminal and C-terminal ions to restrain false positives, but also establishes frequency matrix for post-translational modifications to speculate unknown type of post-translational modifications.Compared with Spectral Alignment, Modified Spectral Alignment not only improves the correctness by 30 percent, but also improves the capability of tolerating noises.