| As more and more gene (genome) have been sequenced, biological data has increased quantitatively, there are also more and more organizations and databases which are founded to store and maintain those data. Our principal purposes are analyzing and researching biological data, however, we found us stuck by many troubles caused by serious diversity of those kinds and storage ways of biological data. So how to store biological data in a unified and valid way became a big burning issue. Meanwhile, as we have stepped in the post-genome era, it is inevitable to break the traditional storage ways. A new way will provide us much convinces when we try to obtain valuable information more efficiently.The main problems of biological data storage are semantic, syntactic and schema heterogeneous. For solving these problems better, we brought XML into this dissertation as a data format and a medium of implementation of the storage patterns considering its current application in biological data storage. Meanwhile we introduced its concepts of object-oriented into our research, according to the characteristics of the storage patterns. In that case, there is only one unified type of information organization during the whole process of biological data storage, which is from information abstraction to data generation.We analyzed, abstracted and classified the bio-information from different angles according to the data's own characteristics. As a result, we got the basic structure of biological data. Then based on this structure, we proposed four bio-information storage patterns according to different kinds of storage requires.Storage patterns are based on the concepts of software design patterns, use object-oriented method and XML to implement the modularization and unionization of biological data storage. Meanwhile, these storage patterns not only correspond with the characteristic of bio-information storage and research, also can lay a good foundation for the future research by their own reusability. |
PDF Full Text Request