Annotation Call July 8, 2014
SGD: Rama, Edith, Stacia
MGI: David, Judy, Li, Mary
Berkeley: Suzi, Chris, Mono, Heiko
EBI: Rachael, Paola, Prudence, Jane
New QC checks
- Reference field
For each annotation, should there be either a PMID or a GO_REF as reference, with option to have an internal MOD reference in addition? i.e. it shouldn't be allowed to have an internal reference alone. There doesn't appear to be a hard check on this.
E.g. Gramene have thousands of annotations using an RCA evidence code and an internal Gramene reference, e.g.
GR_protein B8A749 OsI_01599 GO:0005739 GR_REF:7047 RCA C Putative uncharacterized protein protein taxon:39946 20090810 GR
The reference is detailed here; http://archive.gramene.org/db/literature/pub_search?ref_id=7047
GR_protein Q947Z9 OSJNBa0050M22.8 GO:0004252 GR_REF:8030 RCA InterPro:IPR000209 F Putative retroelement protein taxon:39947 20090212 GR
- RCA evidence
1 year shelf life for Annotations made with RCA. Groups should update the date or remove the annotations after the one year limit.
Source #_of_old_RCA_annotations gramene_oryza 120035 tair 95741 dictyBase 2513 aspgd 1538 pombase 484 GeneDB_Pfalciparum 7003 PAMGO_Mgrisea 21560 ecocyc 171 goa_pdb 141 mgi 105 GeneDB_Lmajor 50 goa_human 35 goa_ref_human 35 fb 19 GeneDB_Tbrucei 14 wb 10 goa_chicken 6 goa_ref_chicken 4 rgd 3 sgd 1
The new website went live last Thur. We encourage all of you to get accounts and get familiar with the site, how to edit etc.
URL to edit: http://www.geneontology.org/user/login
Help page on how to edit/format: http://wiki.geneontology.org/index.php/Editing_GO_Site
It is difficult to track references if just the MOD internal reference is used (if there are typos then the reference did not mean anything). Many groups use abstracts to annotate which do not have a PMID or GO_Ref. Rachael will check to see how many of these are there in the annotations. (Post meeting she sent the breakdown of reference #s for various groups, File:References GOC Files.xlsx)
How beneficial are these RCA annotations? Have we compared them with IEAs to see what we are loosing? Li mentioned that MGI has RCA annotations from individual papers which are still valid. Suggestion is to look at those and update the date for those annotations. But if there are more than a handful of those is it worth the time to review the papers and update the date. This is something each group should decide how to implement (groups can update the date programmatically too).
Consortium members are urged to get an account and edit away. Judy requested Ruth to post her newsletters on the new site.