yrGATE: a web-based gene-structure annotation tool for the identification and dissemination of eukaryotic genes

Complete and accurate gene structure annotation is a prerequisite for the success of many types of genomic projects. For example, gene expression studies based on gene probes would be misleading unless the gene probes uniquely labelled distinct genes. Identification of potential transcription signals relies on correct determination of transcriptional start and termination sites. Characterization of orthologs or paralogs and other studies of molecular phylogeny are also compromised by incomplete or inaccurate gene structure annotation.Gene structure determination is particularly difficult for eukaryotic genomes. Here, we focus on protein-coding genes. In higher eukaryotes, most of these genes contain introns, and a large fraction of the genes appear to permit alternative splicing [1-3]. High-throughput computational gene structure annotation has been highly successful in providing a first glimpse of the gene content of a genome, but current methods fall short of the goal of complete and accurate gene structure annotation (for example, [4-6]). Recent research has focused on improving prediction sensitivity and specificity by combining multiple sources of evidence [7-9]. However, complexities of transcription and pre-mRNA processing, such as introns in non-coding regions, non-canonical splice sites, and utilization of alternative splice sites, still pose formidable challenges for merely computational methods. Re-annotation efforts for most eukaryotic model genomes have, therefore, relied in large part on manual inspection of gene structure evidence [5,10,11]. However, manual annotation also has shortcomings, such as being typically time-consuming, having exclusive participation, and providing annotations only intermittently [4,10,12].A policy of 'open annotation', using the internet as the forum for annotation, and bringing annotation into the mainstream has been suggested as a means to eliminate the restraints of manual annotation and to develop high quality gene anno