|
Processing numerous sequences for a bioinformatics problem, specifically for time
consuming computations substantially benefits from XGrid that harnesses Mac networks.
High performance is achieved with the help from a large number of machines that
execute jobs in parallel; XGrid organizes the process. We will describe our development
and show scripts used for a controller (called the GridServer), agent computers
(called GridAgents), and clients; three components needed in an XGrid solution.
Our solution employs a network of more than 200 Mac computers, each running a A/G
Blast computation. We also discuss statistics that represent speedups in processing
sequences for one of our bioinformatics projects including A/T Blast and InterPro.
|