Vertebrate Annotation

Human Genetics

Archive Page

This page is maintained as a historical record and is no longer being updated.

The HAVANA team relocated to EMBL-EBI in 2017 and continues to create reference gene annotation as part of Ensembl, within Paul Flicek’s team. http://www.ebi.ac.uk/research/flicek

The value of a genome is only as good as its annotation. To create a gold standard reference annotation the Human and Vertebrate Analysis and Annotation (HAVANA) team uses tools developed in-house to manually annotate human, mouse, zebrafish and other vertebrate genomes. This annotation appears in the the Vega browser.

The Sanger Institute has made large contributions to a large number of vertebrate genome sequences, including all or part of human chromosomes 1, 6, 9, 10, 13, 20, 22 and X and mouse chromosomes 2, 4, 11 and X, and the full Danio rerio (zebrafish) genome sequence. The Institute has also sequenced or continues to sequence selected parts of other vertebrate genomes, including candidate diabetes gene regions (in reference and non-obese diabetic (NOD) mouse strains) and MHC regions (in wallaby, Tasmanian devil, gorilla, dog, pig, human haplotypes and mouse strains). The HAVANA team provides the manual annotation for these and other genome sequences.

The HAVANA group puts special emphasis on splice variants and pseudogenes, two areas still underdeveloped in automated annotation systems, as well as poly-adenylation features. Also, where other systems concentrate on, or are limited to, protein-coding genes, many HAVANA transcripts are annotated without a protein-coding region. These transcripts may function as non-coding RNAs or they may be incomplete gene fragments for which the coding sequence cannot yet be determined.

The HAVANA group requires that all annotated gene structures (transcripts) are supported by transcriptional evidence, either from cDNA, EST or protein sequences. As such not all annotated transcripts are necessarily complete. Support does not need to come from locus-specific evidence, but can also be homologous, paralogous or orthologous.

While the transcript and protein sequences are the most important pieces of information, HAVANA annotation takes into account and uses other data, such as CpG islands, gene predictions, repeats and genome signatures. Because the annotation software used is DAS (Distributed Annotation System) aware, the HAVANA team can link to external data sources. Ensembl gene models and data from GENCODE collaborators are some of the DAS sources the HAVANA group uses. HAVANA sources are under constant review and subject change. For example, the group recently started to use data from new technologies such as RNAseq and protein mass spectrometry in its annotation efforts.

The team aims to develop accurate and comprehensive annotation representing the full complexity of gene loci and their features. Manual annotation is especially important in areas that are not well catered for by automated annotation systems, such as splice variation, pseudogenes, conserved gene families, duplications and non-coding genes. The HAVANA team constantly updates its methods by incorporating new data sources that are created as new technologies are developed. HAVANA annotation is freely available through genome browsers, including VEGA, Ensembl and UCSC.

If you have any queries regarding our annotation, please contact us at havana-help@sanger.ac.uk.