A high-throughput distributed DNA sequence analysis and database system

The National Center for Genome Resources (NCGR) has developed a high-throughput DNA (deoxyribonucleic acid) sequence analysis pipeline, which allows researchers at remote sites to submit biological sequence information for rapid analysis, the results of which can be queried through a Web interface. Behind the browser interface is a relational database used to manage both the raw data and the results of the different analyses performed, and a server, which performs those analyses. The system allows multiple contributors to submit data and also allows the data to be marked as "private" or as available to the general public. The CPU-intensive part of the processing is done on a 40- processor domain of a Sun Enterprise 10000 computer, which is represented by a distributed system of software objects, implemented in CORBATM (Common Object Request Broker ArchitectureTM). In this paper we discuss the architecture of the pipeline, the database support, types of DNA sequence analysis used, the distributed analysis system, and the capabilities of the Web interface. As a case study, we present data from an ongoing collaborative project in which expressed sequence tags (ESTs) from Medicago truncatula are being processed. M. truncatula is a plant that is used as a research model for crops in the legume family, an economically important group of food and forage plants.