Exploring prokaryotic taxonomy

Abstract

Techniques drawn from exploratory data analysis, using tools found in the S-Plus statistical software package, have been used to inspect and maintain theBergey's Taxonomic Outlineand to move towards an automated and community-based means of working on the outline. These techniques can be used to classify sequences from unnamed and uncultured organisms, to visualize errors in the taxonomy or in the curation of the sequences, to suggest emendations to the taxonomy or to the classification of extant species and to complement the visualization of phylogenies based on treeing methods. A dataset of more than 9200 aligned small-subunit rRNA sequences was analysed in the context of the current taxonomic outline. The use of the algorithm in exploring and modifying the taxonomy is illustrated with an example drawn from the familyComamonadaceae.