An official website of the United States government.
What is manual curation?
Manual curation in the context of a genome project is when an individual or community manually edits features on the genome, which are often computationally predicted. Both the structure of the predicted feature can be edited, as well as the function of the feature, or the feature metadata. More and more, non-model genomes are curated by a community of experts and volunteers; this is called ‘community curation’. Leaders or organizers of a community curation effort are called 'community contacts'.
Why manually curate a genome?
“Incorrect annotations poison every experiment that makes use of them.” Yandell and Ence 2012.
Manual curation of features on a genome is a critical step in assessing and improving gene prediction accuracy. Additionally, manual curation of genes and gene families by experts significantly enhances the value of gene predictions for the entire research community.
The i5k Workspace@NAL offers the Web Apollo software tool for groups interested in manually curating features on their genome assembly. We are happy to consult with you if you are interested in starting a community curation project of your own - contact us for more information.
Information for community contacts and genome coordinators:
- If you have a genome assembly that you would like to manually curate, please read our guidelines on starting the manual curation process for your genome.
- Contact information for other community contacts is available via our organism pages.
Information for manual curators:
- A list of our Jbrowse/Web Apollo instances.
- Register to annotate one of the genome assemblies that we host.
Our guidelines and tutorials:
- Annotation rules for the i5k pilot project
- A manual curation example
- Information on the Replaced Models field (for projects creating an official gene set)
Other useful links and tutorials:
- The Web Apollo user guide
- Monica Munoz-Torres' presentations on Web Apollo
- A useful guide to gene annotation
- FlyBase's rules for gene model annotation