Automating dChip: toward reproducible sharing of microarray data analysis

<p>Abstract</p> <p>Background</p> <p>During the past decade, many software packages have been developed for analysis and visualization of various types of microarrays. We have developed and maintained the widely used dChip as a microarray analysis software package acces...

Full description

Bibliographic Details
Main Author: Li Cheng
Format: Article
Language:English
Published: BMC 2008-05-01
Series:BMC Bioinformatics
Online Access:http://www.biomedcentral.com/1471-2105/9/231
Description
Summary:<p>Abstract</p> <p>Background</p> <p>During the past decade, many software packages have been developed for analysis and visualization of various types of microarrays. We have developed and maintained the widely used dChip as a microarray analysis software package accessible to both biologist and data analysts. However, challenges arise when dChip users want to analyze large number of arrays automatically and share data analysis procedures and parameters. Improvement is also needed when the dChip user support team tries to identify the causes of reported analysis errors or bugs from users.</p> <p>Results</p> <p>We report here implementation and application of the dChip automation module. Through this module, dChip automation files can be created to include menu steps, parameters, and data viewpoints to run automatically. A data-packaging function allows convenient transfer from one user to another of the dChip software, microarray data, and analysis procedures, so that the second user can reproduce the entire analysis session of the first user. An analysis report file can also be generated during an automated run, including analysis logs, user comments, and viewpoint screenshots.</p> <p>Conclusion</p> <p>The dChip automation module is a step toward reproducible research, and it can prompt a more convenient and reproducible mechanism for sharing microarray software, data, and analysis procedures and results. Automation data packages can also be used as publication supplements. Similar automation mechanisms could be valuable to the research community if implemented in other genomics and bioinformatics software packages.</p>
ISSN:1471-2105