Other taxa annotations (Retired)
Sending a gene association file to GOA
GOA will only take manual, non-ISS annotations that have been made to protein UniProt accession numbers. Files will be re-integrated into the GOA dataset each month, so any additions, changes or deletions made to the file will be displayed after each GOA release. Such imported annotations will, of course, acknowledge the group that the annotations originated from (using the database name provided to us in the assigned_by column), and the annotations will remain the responsibility of the submitting database - any changes that need to be made to the annotations will be communicated back to the group responsible.
As promised, here is the URL for MGI's non-mouse gene association file that GOA now integrates from each month:
Annotations should supplied in the gene association file format (15 column tab-delimited). Although columns 3(db_object_symbol), 9(aspect), 10(db_object_name), 11(synonym) and 12(db_object_type) do not need to be filled in as GOA will obtain this information from our own pipelines.
MGI method for providing annotations to GOA
- When curators are creating ISS annotations for mouse genes, they record the PMID, the evidence code, and any values for the with field as if they were making the annotation for the gene product where the experiment is being done. This info is included in our structured notes.The protein that is used in the experiment is the one that appears in the 'with' field of the mouse ISS annotation.
- We have a script that mines these UniProt IDs and the info in the structured notes and creates a gene association file that goes on our ftp site:
- Daniel harvests this file, adds in the additional info that is required from the UniProt records and incorporates these annotations into UniProt.
- GOA sends MGI an e-mail about any annotations that don't pass a QC check.