Genome BioInformatics Research Lab


  ENCODE logo

GENCODE main page

The GENCODE Pipeline

GENCODE FTP directory

GENCODE annotations

Submitted predictions

GENCODE/E-GASP Genome Browser

ENCODE Gene Prediction Workshop - EGASP/2005


The GENCODE group within the ENCODE project, and the HAVANA team at the Sanger Institute have established a collaboration to produce a high quality annotation (with strong manual and experimenal components) of the gene content of all of the ENCODE regions. The ENCODE Gene Prediction Workshop has been organized to coincide with the finalization of the annotation in the ENCODE regions.

The workshop has two main goals:

1. to evaluate how well automatic methods are able to reproduce the (costly and time consuming) manual and experimental gene annotation of the human genome (the main focus being protein-coding genes),
2. to assess how complete is our current knowledge of the gene content of the human genome.

Predictions are evaluated in terms of their ability to reproduce the ENCODE--HAVANA (termed 'GENCODE') annotation, and to predict novel transcripts--not in such annotation. Promising predictions corresponding to novel genes will be validated by RT-PCR.


The high quality annotation of 13 of the ENCODE regions was released at the end of 2004 ("training set"). Participant groups had access to this annotation, as well as, obviously, to all public data on all of the ENCODE regions. Participants were asked to submit their predictions on the remaining 31 ENCODE regions, using whatever methods and data available to them.
An analysis workshop was held at the Wellcome Trust Conference Center in Hinxton, UK, on May 6-7 2005.

Types of predictions

1. methods using any type of available information to reproduce the HAVANA-ENCODE annotation
2. ab initio single-genome methods
3. EST-, mRNA-, and protein-based methods
4. dual- or multiple- genome based methods
5. methods predicting novel genes (in all 44 regions)
6. methods predicting unusual genes (non-canonical splicing, short inronless genes, etc.)
7. exon-only predictions

Further reading

NEW   We are pleased to announce that the EGASP Genome Biology supplement is now available via AMAZON.   NEW

* Presentations (format: powerpoint/pdf) given by E-GASP participants during the analysis workshop.
* Nature News article about E-GASP (Nature 435, 134 (12 May 2005) | doi: 10.1038/435134a).
* EGASP: collaboration through competition to find human genes (Commentary by Guigo & Reese, Nature Methods 2, 575 - 577 (2005) | doi:10.1038/nmeth0805-575).

Organising Committee

Jennifer Ashurst   Wellcome Trust Sanger Institute
Ewan Birney European Bionformatics Institute
Peter Good National Human Genome Research Institute
Roderic Guigó Institut Municipal d'Investigació Mèdica
Tim Hubbard Wellcome Trust Sanger Institute

Advisory Committee

Michael Ashburner   Cambridge University
Vladimir B. Bajic Institute for Infocomm Research
Tom Gingeras Affymetrix
Suzanna Lewis Berkeley
Martin Reese Omicia

  Disclaimer GENCODE webmaster