- 1 Agenda
- 1.1 Welcome to Melanie Courtot
- 1.2 Adding people to managers list
- 1.3 Update on GitHub Migration
- 1.4 There are not 46k human genes
- 1.5 GAF2.1 (addition of pipe/comma to With/From column) (Rama)
- 1.6 Annotating Complexes with GO
- 1.7 Author feedback form
- 1.8 Master GAF file
- 1.9 No more merge ?
- 2 Discussion
Welcome to Melanie Courtot
- She is the new GO project lead at UniProt
Adding people to managers list
- Make sure Moni and Melanie or on the list
Update on GitHub Migration
- https://trello.com/c/nf1HDHwv/149-migrate-ontology-from-svn-sourceforge-to-github (Chris)
- This has become high priority
There are not 46k human genes
This is consistently confusing users and even us. Need to fix:
Human Refernece Proteome Set- http://www.uniprot.org/proteomes/UP000005640 ftp://ftp.ebi.ac.uk/pub/databases/reference_proteomes/
GAF2.1 (addition of pipe/comma to With/From column) (Rama)
- I have put together some specs to make this happen.
- Documentation on the GOC website for GAF 2.1- http://geneontology.org/page/go-annotation-file-gaf-format-21
- Annotation guidelines for use of pipe and comma is pending
- test file for checking scripts is available in github
- protein2GO is almost ready to make this option public
- We want to announce this version on July 1st.
Are there any concerns?
Annotating Complexes with GO
IntAct folks are making GO annotations for complexes and they have sent us some questions about how to annotate tricky situations etc. Myself and Kimberly have gone through the proposal. We recommend forming a working group to discuss the issues and to come up with guidelines. Is Pro making GO annotations to complexes? Any thoughts?
Author feedback form
- We have been asked several times about the quality of our annotations
- Paul T mentioned that, at the Synapse meeting, an author said the annotations made from his paper were not accurate.
- How about we send a form to authors and ask if the annotations made from their papers are correct or not?
Master GAF file
- Many of our users who want all the annotations for all taxons grab the UniProt file. We should be providing a master GAF file with all annotations
No more merge ?
(Pascale email; see minutes from Ontology call http://wiki.geneontology.org/index.php/Ontology_meeting_2015-05-28
Moni and Melanie
They both are subscribed to the go-managers list. Paola and others at EBI are helping Melanie get familiar with GOhelp and everything GO. Melanie is the GO editorial lead in Helen Parkinson's group and the GO annotation project lead in Claire's group with collaboration with Maria Martin's group. The funding has come from EMBL-EBI, not UniProt.
- SF future no looking good. Might go down sooner than we think. Chris is making regular backups of SF and has a script to migrate the tickets over to github. Once he gets the github usernames mapped tickets will be transferred. He will make an announcement (email, annotation call etc) when we switch.
- Rama: there are some outside users who have tickets in SF. How do we handle thos? Chris: Can't be handled. Hope they contact us via go-helpdesk
- Lot of issues with 46K human genes. messes up term enrichment because 2 versions of the protein is represented. This is because annotations are sometimes made to non-reference protein IDs. In p2GO if you query by the gene name you get multiple IDs (including the unreviewed TremBL IDs) and curators pick one of those trembl ids. Can we make p2GO show only one ID when we search for a gene name?
- Judy- Ed Marcote's recent paper clearly shows how duplicate genes in human have different functions. We might miss some annotations if we annotate to the isoform ID. We have to be careful.
- we should make sure we have a similar process for all genomes not just humans
- Kimberly- in worm there is evidence for 2 isoforms based on ESTs. No protein evidence. how do we resolve this? Do groups have resoures to merge records?
- Form a working group to figure this out.
- Most of the GOC members did not see this paper before it was published. Groups that are publishing a paper with GO data should circulate the paper among the consortium members, minimally PIs to make sure the message about GO is accurate/consistent.
- No flags raised. We can go ahead and make an announcement. AmiGO will continue to display everything as pipes.
- Melanie- are there going to be scripts to convert between formats? CHris- Yes!
- will form a small working group to work out the details. Judy and HArold will join this group as PRO representatives.
Master GAF file
- Everybody agreed to this idea. Rama has made a github ticket already for this item
No More merge
- this was implemented more for streamlining Ontology development. If groups use a script to automatically transfer annotations they have to accommodate this change. Will mention this in the Annotation call.
No time to discuss!