ABSTRACT
Besides the obvious reason that automation provides eªciency for custom analysis of genome-scale data sets, there are several reasons for adopting work£ows in bioinformatics. First, a formal work£ow can be used to document the analysis process used to derive the annotation results. ™is document o©en provides useful information for the “Methods” section when the results are eventually published. ™ere is a growing trend to include
formal work£ows as supplementary information for papers, because they are communicative and provide the background for the reproducibility of results, which can be helpful for readers and reviewers alike. Work£ows can be rerun periodically to ensure that the genome annotation results are up to date, with respect to the reference databases they were compared to, and by extension the current scienti¥c knowledge base. Creating a work-£ow can save e¤ort down the road through its reuse. For example, through slight modi¥cations it might be used to perform di¤erent but related analyses in follow-up research.
