WASABI: An Automated Sequence Processing System for Multigene Phylogenies

Abstract

WASABI (Web Accessible Sequence Analysis for Biological Inference) is a software that provides the computational infrastructure for multiuser multigene phylogenetic projects. In addition to serving as a web accessible sequence and voucher database, WASABI pro vides the functionality for automated sequence analysis, verification, and data set assembly WASABI is unique as it supports the user by carrying out all steps from processing raw sequence data to a multigene sequence analysis within a strong phylogenetic context, and thus differs from tools for databasing and analyzing finalized gene sequences (e.g., ARB [Ludwig et al., 2004], Tax Man [www.nematodes.org/bioinformatics/Taxman/ index.shtml]) or for the processing of ESTs (e.g., Stack Pack [www.egenetics.com/stackpack.html], PartiGene [Parkinson et al., 2004]). Although originally developed as a computational framework for the Assembling the Fungal Tree of Life (AFTOL) project (www.aftol.org, www.lutzonilab.net/aftol), part of the NSF-funded As sembling the Tree of Life initiative (ATOL, atol.sdsc.edu), WASABI can easily be adapted to the specific needs of most multiuser multigene sequencing projects. Currently, components of WASABI are adapted for three other Tree of Life projects: Assembling the Beetle Tree of Life (BToL, insects.oeb.harvard.edu/ATOL/ people.htm), Cnidaria AToL (CnidTol, cnidar ian.info), and Assembling the Tree of Eukaryotic Diversity (Eu-Tree, www.biology.uiowa.edu/eu_tree). WASABI should not be confused with the identically named Web Application for the Semantic Architec ture of Biodiversity Informatics (Perry and Vieglais, 2006). The AFTOL project, which was initiated in January 2003, is a collaboration centered upon five laboratories in four U.S. universities, with the international partici pation of more than 100 scientists. The aim of AFTOL is to improve our knowledge of fungal phylogeny by col lecting sequence data from eight genetic loci for 1500 fungal taxa across all major fungal lineages, as well as developmental and ultrastructural data from a re duced set of selected taxa (Spatafora, 2005). AFTOL participants can contribute to this project by providing fresh specimens, culture strains, frozen material, DNA samples, primary sequence data for specific genes, or phenotypic data for targeted taxa. All information, in cluding voucher information and the availability and lo cation of material/samples, is attached to a DNA sample with a specific AFTOL number. These data are stored in a central database, managed by and accessed through WASABI (www.lutzonilab.net/aftol). It is generally agreed that nucleotide sequence data derived from multiple unlinked loci are essential to re solve, with high confidence, relationships among a large number of species and ultimately assemble the tree of life (Lutzoni et al., 2004; James et al., 2006). Two main opera tions are required for large-scale phylogenetic studies? data acquisition and phylogenetic analysis. We define here the data acquisition process as all steps from the collection of samples in the field to the generation of con catenated data matrices of multiple data partitions (phe notypic and genomic) for which each data partition was tested for congruence until no significant conflicts were detected among data partitions; i.e., until the data are ready for final phylogenetic analyses. Most research ef forts have been concentrated on improving phylogenetic analyses, with the resulting situation that data acquisi tion remains a time-consuming, manual, and error-prone process for large-scale multipartition phylogenetic stud ies. The limitations of this practice were revealed in a re cent survey of 595 fungal phylogenetic trees found in 560 articles published from 1990 to 2003 (Lutzoni et al, 2004). Despite the enormous progress in the fields of genomics and information technology, as well as new theoretical and technical innovations derived from multidisci