Virtual Supercomputer Using Volunteer Computing

1. Introduction To Next Generation Sequencing

Modern quantitative biology has changed the perspective of data rich genomic sequencing technology. Large scale genomic data analysis requires the need for a new computational framework supported by High Performance Computing. One such application is the Next Generation Sequencing (NGS), which deals with terabytes or petabytes of genome data requiring high computational power.

Next Generation Sequencing (NGS) (Wilson et al., 2002; Narzisi et al., 2011) is a technique of sequencing the exact order of nucleotides which form the basic building blocks of Deoxyribonucleic Acid (DNA). NGS with a market size of over 2.7 billion dollars has diverse uses in fields of biological sciences ranging from identification of diseases in human beings to invention of sequence for novel species. Traditionally sequencing was done by treating DNA chemically and identifying nucleotides using color codes, but this technique of sequencing is not suitable for organisms with just thousands of nucleotides. Earlier, the cost of producing base pair information stored as 'reads' was limited to wet laboratory techniques and was very expensive. Hence the rate of production of data was very slow, but new sequencing technologies combined with wet lab techniques and information technology started producing millions to billions of short ‘reads’ quickly. The traditional assembly tools used earlier was incapable to handle this huge data.

To overcome these problems a number of assembly technologies have been invented that uses computations performed by computer, also known as the In silico approach. These assemblers started with small datasets and were effective. As the size of 'reads' increased, the assemblers required either a single computer with very large amounts of memory and computing resources or the data to be sent to third party for execution such as cloud computing which might lead to security concerns. These constraints make the analysis of huge amount of genomic data a tedious task.

An alternate solution to Cloud and Hadoop is to use volunteer computing which is proposed and explained in this chapter. In particular emphasis is on recommending a solution to Next Generation Sequencing (NGS) which uses an open source grid middleware namely Berkeley Open Infrastructure for Network Computing (BOINC) designed to handle various applications that require high computational power, data storage or both. This will be a great enabler for bioinformatics scientists to create applications that use public computing resources.