The Personal Biological Information Database Based On Gene Ontology And Its Application On The Prediction Of Protein Subcellular Location

The research and development of the advanced molecular biological information technology has become the hot point and the pioneering area of the international scientific reaseach. The accumulation of a large amount of sequence data got from the biological experiment forms hundreds of biological information database. Those databases collect and process the biological experimental sequence data according to some specific goal, and provide the service of data search and processing.The biological information database has become the indispensable research tool in the advanced biology. In the advanced biological research, the computer has the same importance with the experiment equipement and is especially needed to analyze a variey of DNA sequences rapidly.The personal database is different with the ordinary database. It is the professional database in some research field used by some people for some use. In this passage, the professional biological information database used on the biological information research, which is called the personal biological information database, is discussed. As for those individual databases and those research groups who have the right to let their personal information be open to the public, their research results can be published in time or stored into the personal database developed by themselves or other personal databases.The author developes a biological information personal database system which is used on the research of the prediction of the proteinsubcellular location. The research content in this passage includes the following aspects:1) We describe the feature of Gene Ontology and make the design of the biological information databases. A personal biological information database system based on GO(BIODB_GO) is also developed in this passage. The author proposes to annotate the protein sequence using GO and search the protein sequences using GO terms, which can exactly find the needed sequence and reduce the redundancy and insufficiency of the final result of searching the sequence data.2) After generalizing the methods of the prediction approach of the protein's subcellular location, the author brings forward a new stochastic signal processing approach with all sequence feature. The three non-linear parameters (Lyapunov Index, Besseli Function, Chepshev Filter) are also used as the algrithom parameters used in predicting the protein's subcelluar location.3) As for the stochastic signal processing approach with all sequence feature, the system weight-value is decided finally through the feedback adjustment method and a higher prediction result is got finally.4) The author applies the BIODB_GO system on the prediction of protein's subcellular location. The MATLAB is used to realize the new stochastic signal processing approach with the complete-sequence feature. The realized program is integrated into the BIODBGO system. The sequence data used by the program can be taken from the sequence data searched by the GO research method, so that the analyze of the protein sequence can be done quickly.The BIODB_GO provides a good research platform for our group's biological information research. Moreover, the system can seta solid basement for the better development of the intelligent biological information database system.