Ensembl variation resources.
Hunt SE, McLaren W, Gil L, Thormann A, Schuilenburg H, Sheppard D, Parton A, Armean IM, Trevanion SJ, Flicek P, Cunningham F. · 2018
Read Ensembl variation resources. by Hunt SE, McLaren W, Gil L, Thormann A, Schuilenburg H, Sheppard D, Parton A, Armean IM, Trevanion SJ, Flicek P, Cunningham F. (2018) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc6310513.txt.
Introduction
Hunt and co-authors describe the Ensembl variation resources for genome data. The article appeared in Database, the journal of biological databases. Linking DNA variation to traits and disease is a core goal of genomics. Data on variants sit scattered across many sources. Ensembl gathers them into one consistent system.
The team builds pipelines that compare variants with genome annotations. Data are harmonized and served in standard formats. Users can browse them visually or query them by code. Sources cover allele frequencies and disease links. Predicted effects on genes are included too. All tools and data are public. The resource helps labs read a person's genome with context.
Catalog record
| CREATOR | Hunt SE, McLaren W, Gil L, Thormann A, Schuilenburg H, Sheppard D, Parton A, Armean IM, Trevanion SJ, Flicek P, Cunningham F. |
|---|---|
| YEAR | 2018 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 43 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Ensembl variation resources By Sarah E Hunt, William McLaren, Laurent Gil, Anja Thormann, Helen Schuilenburg, Dan Sheppard, Andrew Parton, Irina M Armean, Stephen J Trevanion, Paul Flicek, Fiona Cunningham Published in: Database: The Journal of Biological Databases and Curation DOI: 10.1093/database/bay119 ABSTRACT Abstract The major goal of sequencing humans and many other species is to understand the link between genomic variation, phenotype and disease. There are numerous valuable and well-established variation resources, but collating and making sense of non-homogeneous, often large-scale data sets from disparate sources remains a challenge. Without a systematic catalogue of these data and appropriate query and annotation tools, understanding the genome sequence of an individual and assessing their disease risk is impossible. In Ensembl, we substantially solve this problem: we develop methods to facilitate data integration and broad access; aggregate information in a consistent manner and make it available a variety of standard formats, both visually and programmatically; build analysis pipelines to compare variants to comprehensive genomic annotation sets; and make all tools and data publicly available. INTRODUCTION One of the most important challenges in contemporary biomedical research is to understand how changes in genome sequences lead to different phenotypes and influence disease. Collecting and interpreting genomic sequence variation is the key to addressing these and other questions. Indeed, recent results have provided insight into disease susceptibility (1), treatment response (2), development of desirable traits in animals (3) and crop plants (4) and understanding population genetics in many species (5). Human genetics study designs in particular are transitioning from targeting a handful of candidate genes to full exome or genome sequencing to find trait-associated loci in individuals, families or populations (6, 7). In all study designs, sites of genome variation alone are insufficient to draw conclusions. To interpret results, it is essential to have access to additional data such as allele frequencies in reference populations, reported disease associations and predicted functional impact on genes (8). To facilitate this important and growing area of research, Ensembl identifies, integrates and organizes comparable data from different sources into a system that can be queried and browsed using uniform access and analysis methods. Here, we describe how the variation and phenotype data sources incorporated into Ensembl are harmonized, visualized and programmatically accessed. We focus on data sources relevant for genome interpretation, including for understanding human phenotypes and disease. Major changes since our last report [9, that described the variation resources present in Ensembl release 56 (September 2009)] include the incorporation of more extensive phenotype and disease annotation in multiple species, the inclusion of variation citation data, improved allele normalisation and equivalence algorithms and infrastructure changes to support the presentation and visualisation of many millions of genotypes. [Figure: A variant summary view for rs148698650 that displays the global minor allele frequency from the 1000 Genomes Project, other variants at the same location and links to projects providing additional information about the variant. URL: https://www.ensembl.org/Homo_sapiens/Variation/Explore?v=rs148698650] DATA ACCESS The genome browser For analysis of small areas of the genome, such as variation in a single gene or transcript, visual displays remain the key to explore, analyse and communicate scientific findings. We provide access to the data we incorporate through several different interactive displays via the Ensembl genome browser (www.ensembl.org). From the Ensembl home page, searching for genomic regions, genes, phenotypes, diseases or variant identifiers provides access to different data views. For example, variant-specific web pages (see Figure 1) are a convenient way to visualize data aggregated from a wide range of different variant-focused projects combined with the results of our analyses. The information is presented both as a summary and as a set of detailed graphical views and tables. These are described in more detail below. The Ensembl genome browser also provides displays of genomic regions in which variants are presented alongside transcripts, regulatory features and conservation information. Variants are coloured by the most severe predicted impact they have on gene function. This ‘Region in detail’ view (see Figure 2) can be configured to display tracks of specific variants, such as those from ClinVar, those with phenotype annotations or those assayed on selected genotyping arrays, such as the Illumina ImmunoChip (10) and the Affymetrix Chicken600K array (11). [Figure: The ‘Region in detail’ view showing 5 of the available 66 tracks of variant data. URL: https://www.ensembl.org/Homo_sapiens/Location/View?r=10:79069035-79316528] Our dedicated gene pages include tables describing all variants in the gene region with many filtering options including by consequence or variant type. These are configurable to improve readability and analysis capabilities, especially in genes with a large quantity of variant data. We also provide configurable phenotype pages that show variants and genes associated with a phenotype, trait or disease. On both pages, data are imported from multiple sources and presented in a single table for easy comparison. Application programming interfaces To facilitate direct data access for flexible analysis and to support the web displays described above, we have developed and extended our methods for rapid bulk data retrieval. Custom queries can be written using our mature and comprehensive Perl application programming interface (API), for which extensive documentation and detailed tutorials are available on the Ensembl website. We have also implemented a Representational State Transfer (REST) API (12) for language agnostic data access. Together, these enable convenient integration of our tools and data into multiple other websites and analysis pipelines. Variant Effect Predictor The data assembled in the Ensembl databases can be used to annotate variants and predict their functional impact using the Ensembl Variant Effect Predictor (VEP) tool (13). VEP provides a simple, yet powerful, interface using the Ensembl API, which enables the user to annotate an entire human genome (around 4 million variants) in less than an hour and an exome (around 200 000 variants) in less than 5 minutes. It can be accessed via a web interface, a stand-alone script and a REST API. The web interface is integrated with other Ensembl views allowing navigation directly from VEP results to detailed information on any transcripts or to previously known variants that match an input variant. The stand-alone script is highly configurable and can be extended to incorporate additional data from Ensembl (using the Perl API) or other sources. For more detailed information, see (13). HETEROGENEOUS DATA INTEGRATION Over the past 8 years, we have increased the number of supported species, the range of data types we hold and volume of data we store, analyse and serve. For example, we have extended our schema to include variant data citation data and greatly improved our management of phenotype and disease annotations to incorporate additional data sources. Short variants The Ensembl variation databases currently contain publicly available data for 20 vertebrate species from sources including dbSNP (14), the European Variation Archive (EVA) (https://www.ebi.ac.uk/eva/) and ClinVar (15). To allow all information for a locus to be considered together, we collate data where possible and index the merged records with identifiers from multiple resources. This enables searching directly with accession numbers from multiple databases including dbSNP, UniProt (16), PharmGKB (17) and ClinVar. Since our last report, the number of short variants held in Ensembl databases has risen from 56 million to over 1166 million as of Ensembl release 93 (July 2018). The distribution of data across species has also changed