Construction of Federated Genome Knowledgebases by Grid Computing

基因组相关数据分别存储在不同方案和格式的数据库中，如DNA核苷酸序列的DDBJ/EMBL/GenBank，蛋白质序列的PIR和Swiss-Prot，以及蛋白质三级结构的PDB。此外，还根据数据的领域或数据库的用途建立了一些数据库。最近，与基因组相关的数据库数量达到700多个数据库。在新的数据库管理方法和新的访问方法出现之前，用户很难找到新的有价值的知识，从这些角度出发，本研究针对以下主题：(1)基于本体的不同科学领域的统一数据访问，数据类型多样，属于不同的领域，即DNA核苷酸、蛋白质和细胞成分。因此，很难将这些数据库统一为单一共同方案。因此，我们开发了本体及其可用性技术，作为每个数据领域或跨学科领域的公共描述。(2)为联邦数据库的查询处理进行组合爆炸是可能的。为了解决这个问题，我们开发了一种定制网格环境的技术。

The genome-related data have been separately stored in different databases of which schemes and formats are different ; such as DDBJ/EMBL/GenBank for DNA nucleotide sequences, PIR and SWISS-PROT for protein sequences, and PDB for protein tertiary structures. Moreover, a number of databases have been constructed according to the domain of data or the purpose of databases. Recently, the number of genome-related databases reaches more than 700 databases. It is difficult for users to find new and valuable knowledge until new database management methods and new access methods are developed.From these viewpoints, this research aims at the following themes :(1)Ontology-based unified data access in different scientific domains.There are many types of data which belong to different domains ; namely, DNA nucleotides, proteins, and cellular components. For this reason, it is difficult to unify such databases into single common scheme. Thus we have developed ontologies and their usability technology as common descriptions of each data domain or interdisciplinary areas.(2)It is possible to make combinatorial explosion for the query processing of the federated databases. To cope with this issue, we have developed a technology for customizing a grid environment.