Based On The Segment Information Fusion Protein Subcellular Sites And Testing Methods

Bioinformatics is a cross-discipline, which includes all aspects of the biological information such as acquisition, processing, storage, distribution, analysis and interpretation, etc. Bioinformatics uses mathematics, computer science and biology tools, to clarify and understand large amounts of data contained in the biological significance. The research of bioinformatics is very broad, and protein subcellular localization sites is one of the most popular research topics. A complete cell can be divided into many different cell regions (cytoplasmic and chloroplast, for example), and most proteins will be transferred to specific cells after they are synthesized in the ribosome. Proteins only in the appropriate site could perform their specific functions, or it will causes other effects.The main contents of this paper include as follows:In Chapter 1, we introduced the background, research object, andresearch contents of bioinformatics and the work of this paper.In Chapter 2, we introduced some machine learning methods about classification problems in bioinformatics , such as k-Nearest Neighbor, Bayesian Statistics, Neural Network Model and Support Vector Machine, etc.Chapter 3 introduced a new method for predicting protein subcellular location. In this paper, we divided each chain into three parts: N-terminal, middle, and C-terminal. Then, features were extracted from each part and the whole chain independently. These features are amino acid compositions, dipeptides, and stereochemical properties. Finally, features of different parts are combined and the combined features are used as features of the whole chain. By Jackknife test on the NNPSL dataset, the overall accuracies for prokaryotic and eukaryotic proteins achieve 92.1% and 87.8%.