V Uday Kumar Reddy, Rajashree Shettar and Vidya Niranjan
Invention of new computing techniques like cloud and grid computing has reduced the cost of computation by optimal resource sharing. Yet, many applications are not moved completely into these new technologies mainly because of the unwillingness of the scientists to share the data over internet for security reasons. Even though cost of the hardware has been reduced drastically few applications require high processing power to process or analyze huge scientific data. Also due to high cost required to acquire compute resources many of the scientific applications are yet to materialize completely. One such application is next generation sequencing (NGS) which will have to deal with Terabytes of genome data, which will require high computation power. Hence a super computer is required to efficiently process data.
In this paper, the use of Berkeley Open Infrastructure for Network Computing (BOINC) open source grid middleware has been proposed to enable de novo assembly using a cluster of desktop machines in master and volunteer paradigm. The paradigm can be set up in normal computer laboratories which eliminate both the bandwidth and security concerns of using cloud and grid computing methods over the Internet. This paradigm creates a virtual super computer in laboratories to process data.
