TreeBASE v. 2: A Database of Phylogenetic Knowledge

Abstract

TreeBASE (http://www.treebase.org) is a prototype web-based digital repository of published phylogenetic trees, phylogenetic characters, and sequence alignments. This year we are releasing a new ver- sion, TreeBASE v. 2, that is completely rebuilt in a tiered Java-based architecture using the Hibernate and Spring frameworks. We are removing a number of limitations in the previous version, including lift- ing limits on the size of trees, parsing and storing branch lengths and clade support, allowing storage of continuous or non-discrete characters, and providing a greatly enhanced ability to describe the parameters and steps used in an analysis. In this new version, phylogenies are hashed to allow topological searching, and taxon labels are mapped against name services provided by uBio and NCBI with the result that we offer an approximate normalization of over 80% of our taxon labels, thus allowing the ability to perform higher classification name searching. Additionally, we extend the granularity of data provenance by providing the ability to attach specimen-specific and gene-specific metadata (e.g. basic Darwin Core elements, specimen voucher numbers, and Genbank accession numbers) to segments of data matrix rows. Primary access to the database will continue with a basic web interface, but in addition we will offer a PhyloWS RESTful API and a periodic SQL data dump.