Ensembl 2009

Abstract

The Ensembl project (http://www.ensembl.org) is a comprehensive genome information system featuring an integrated set of genome annotation, databases, and other information for chordate, selected model organism and disease vector genomes. As of release 51 (November 2008), Ensembl fully supports 45 species, and three additional species have preliminary support. New species in the past year include orangutan and six additional low coverage mammalian genomes. Major additions and improvements to Ensembl since our previous report include a major redesign of our website; generation of multiple genome alignments and ancestral sequences using the new Enredo-Pecan-Ortheus pipeline and development of our software infrastructure, particularly to support the Ensembl Genomes project (http://www.ensemblgenomes.org/).

Figure 1.

Screenshots of the Release 51 Ensembl website illustrating the principles of the new design and some of the new features. The figure shows an example of three of the four classes of display view using human gene SLC24A5 as the context. (A) An example of a location based view, showing a region of the genome around the gene. (B) An example of a gene based view, showing the gene tree. (C) An example of a transcript based view, showing supporting evidence for the transcript model. The three tabs across the top of the page, allow rapid navigation between the three classes of view. The fourth variation tab (data not shown) appears if an individual SNP is selected. For each class, the left hand menu lists the different views available. For the location-based views (A), this includes views of a genome at a range of resolutions and genome sequence based comparative genomic views. For the gene-based views (B), this includes textual information about the gene, views of its local genomic environment, views of the gene in the context of its orthologs and paralog relationships with other genomes in the Ensembl system and views of sequence variation within that population. The transcript based views (C) have views similar to the gene based views, but focused around individual transcript structures with more detail. As well as the overall redesign of the navigation between views, there are substantial improvements to many individual views, based on the much more extensive use of AJAX in the new web-code. Examples are the genetree view (B) which allows nodes to be expanded or collapsed interactively, making the view much more usable for large gene families; the substantially redesigned supporting evidence views (C) and the page configuration options on many views (e.g. A) which are much more intuitive than before and have a much greater range of display options.

Figure 2.

Figure shows a smoothed density plot of the GERP conservation scores (30) calculated from the 9-way EPO mammalian genome alignment corresponding to human chromosome X (Ensembl release 51). Four different types of genomic features are plotted: coding exons (red), non-coding exons (pink), regulatory features (blue) and ancestral repeats (black). A GERP score of 0 indicates no evidence of selective constraint, whereas high GERP scores shows evidence of selective constraint. Non-coding exons include all non-coding positions of protein-coding genes. Regulatory features include all regulatory features defined by the Ensembl regulatory build (‘Gene Associated’, ‘Non-Gene Associated’, ‘Promoter Associated’ and ‘Unclassified’). Ancestral repeats include MER type II transposons only, as defined by RepeatMasker. Conserved features, such as exons and regulatory features, are clearly distinguished from repeats, a good indication of the quality of the EPO alignments.