Research On The Models Of Protein Secondary Structure Prediction

With the approach of post-genome era, proteomics is becoming an important research domain in the life science. Prediction of protein secondary structure (PSSP), with a long historic task, is still a challenge in the research of proteomics at present. Any new breakthrough in this area will be helpful to knowing better the function of protein and the human healthy or paroxysm cell constitutes. What is more, it will be an important assistant to some relevant industries such as medical engineering, agriculture, agbiotech, etc. Some work on the study of PSSP using the methods of bioinformatics is done in this thesis.The main work of this thesis are summarized as follows:1. Prediction of lowly homological protein secondary structure is still a difficult problem up to now. When applying the standard hidden Markov model to predict the protein secondary structure, we must set out from the knowledge according to homological protein secondary structure database. Under lowly homological source, standard hidden Markov model is difficult to do PSSP. Aiming at this kind of blemish, we put forward to the hidden semi-Markov Model. It can be applicable to predict the lowly homological protein structure. Using the AA physical and chemical properties, it can judge the tendency of folds and establish the tendency model of part folds.2. The comparability of the part highly homological proteins is studied, and an expert system designed based on the part highly homolog. In addition, a separation rate of the homological part is given and applied to the model of HSMMs.3. The redundancy in the database of protein structure is analyzed, that the protein secondary structure of the international database error is around 25%. Comparing with the highly rigorous prediction, the systematic error is obviously the main factor to affect the accuracy of the protein structure prediction.