Constructing computational pipelines

Abstract

Many bioinformatics tasks involve creating a computational pipeline from existing software components and algorithms. The traditional approach is to glue components together using scripts written in a programming language such as Perl. However, a new, more powerful approach is emerging that promises to revolutionise the way bioinformaticians create applications from existing components, an approach based on the concept of the scientific workflow. Scientific workflows are created in graphical environments known as workflow management systems. They have many benefits over traditional program scripts, including speed of development, portability, and their suitability for developing complex, distributed applications. This chapter explains how to design and implement bioinformatics workflows using free, Open Source software tools, such as the Tav-erna workflow management system. We also demonstrate how new and existing tools can be deployed as Web services so that they can be easily integrated into novel computational pipelines using the scientific workflow paradigm. © 2008 Humana Press, a part of Springer Science+Business Media, LLC.