Metaprogramming bioinformatics in the postgenomic era

Abstract

The number of bioinformatics programs available is continuously growing, along with the knowledge required to run each individual program. As more programs become available, more complex combinations of these programs are being used by scientists. Workflow engines attempt to remove repetitive procedures from these combinations by saving and executing any group of the programs as one larger "meta-program". The output of one program is automatically directed to the input of another thereby creating a "flow' of "work". The first part of this project is the design and development of an easy to use workflow editor that maintains the usability of the original programs and allows for the relatively simple addition of new programs. The result is a workflow editor that currently allows access to over 170 bioinformatics programs each with a simple, common, and descriptive interface. In the second part of this project a specific workflow application for protein antibody imprinting is examined. Part of the workflow is a program that produces a series of potential alignments for an antibody and a protein, has already been developed. An additional program is developed which uses a greedy search algorithm to select the "best" set of alignments. This alignment selection problem is shown to be NP-Complete. These programs represent a real world example of a bioinformatics workflow.