Summary form only given. Bioinformatics is an important application area for grid computing. The grid computing issues required to tackle current bioinformatics challenges include processing power, large-scale data access and management, security, application integration, data integrity and curation, control/automation/tracking of workflows, data format consistency and resource discovery. In this poster, we describe preliminary steps taken to develop a grid environment to advance bioinformatics research. We developed a system called Grendel, with the aims of providing bioinformatics researchers transparent access to basic computational resources used in their research. Grendel is a platform and language independent Web-services based system for distributed resource management utilising Sun Grid Engine that provides a single entry point for computational tasks while keeping the actual resources transparent to the user. Grendel is developed in Java and deployed using the Tomcat. Client libraries have been developed in Perl and Java to provide access to computation resource exported via Grendel.

MOTIVATION: There are a large number of computational programs freely available to bioinformaticians via a client/server, web-based environment. However, the client interface to these tools (typically an html form page) cannot be customized from the client side as it is created by the service provider. The form page is usually generic enough to cater for a wide range of users. However, this implies that a user cannot set as 'default' advanced program parameters on the form or even customize the interface to his/her specific requirements or preferences. Currently, there is a lack of end-user interface environments that can be modified by the user when accessing computer programs available on a remote server running on an intranet or over the Internet. RESULTS: We have implemented a client/server system called ORBIT (Online Researcher's Bioinformatics Interface Tools) where individual clients can have interfaces created and customized to command-line-driven, server-side programs. Thus, Internet-based interfaces can be tailored to a user's specific bioinformatic needs. As interfaces are created on the client machine independent of the server, there can be different interfaces to the same server-side program to cater for different parameter settings. The interface customization is relatively quick (between 10 and 60 min) and all client interfaces are integrated into a single modular environment which will run on any computer platform supporting Java. The system has been developed to allow for a number of future enhancements and features. ORBIT represents an important advance in the way researchers gain access to bioinformatics tools on the Internet.