Annotation Conf. Call, January 10, 2012

Agenda

* IC to/from ISS annotations

Some curators would like to supply a non-model organism with a full set of high-quality annotations, by transferring all appropriate knowledge from a well manually-curated gene product.

Currently only EXP-evidenced annotations can be propagated to other gene products via the ISS evidence code.

What should curators do when they would like to propagate the information from an IC-evidenced annotation to a range of gene products in a related organism?

Full details IC to/from ISS annotations here

* Planned changes to the IEA GO_REFs cited in the UniProt-GOA file

- New GO_REFs proposed to identify annotations differently created for UniProtKB/Swiss-Prot and UniProtKB/TrEMBL entries from the UniProtKB keywords and UniProtKB subcellular locations vocabulary mappings.

- Proposal to rename of spsl and spkw mapping files to better indicate they are UniProt vocabularies

- Intention to transform some incorrect IEAs supplied by the UniProt-GOA files.

Full details here

---

Minutes

Present: Petra, Pascale, Rama, Julie, Karen, Peter, Chris, Rachael, Emily, Prudence, Yasmin, Becky, Paola, Ruth, Li, Lakshmi, Stan, Shur-Jen, Kimberly

* IC to/from ISS annotations

Summary:

Karen: Raised general concern regarding over-enthusiastic transfer of process terms to orthologs where the process is not appropriate.

Most of the call focused on the option 1: enable IC-evidenced annotations to be transferred via the ISS code

Julie: Concerned that introducing a sub-set of ISS-evidenced annotations might lower the quality of the overall ISS annotation set, and make it more difficult for users to understanding the annotation set.

Emily, Pascale, Ruth: IC-evidenced annotations are of high-quality and extremely conservative. Their inclusion alongside EXP-evidenced annotations should not devalue the ISS annotation set.

Rama: are the counts outlined in the document including the automatically inferred annotations? Emily: no, the counts exclude GOC-inferred annotations.

Emily: enabling curators to make ISS annotations from IC evidence would be of great benefit where a curator would like to transfer annotations to a large number of closely-related species which do not have any experimental information available (e.g. from human to other primates).

Peter: this would be valuable for such users, who can not expect to get EXP-level annotations. While we should be transparent about our curation methods, users need to also be partially responsible for educating themselves about the resource that they use.

Karen: preference for option 1, as IC annotations are such high-quality, as supported by experimental evidence, that we would not want to make subtypes of IC.

Chris: we could not only make a specific ECO code for IC-evidenced ISS annotations, but also include a specific GO evidence code; annotations should preferably not be excluded from the GAF files.

Outcome:

Curators agreed to use option one - enable IC-evidenced annotations to be transferred via the ISS code, however a descriptive evidence code needs to be developed to identify this annotation set.

ACTION: create a descriptive ECO and GO evidence code, and circulate around GO list and annotation call.

* Planned changes to the IEA GO_REFs cited in the UniProt-GOA file

Summary:

Part 1: New GO_REFs proposed to identify annotations differently created for UniProtKB/Swiss-Prot and UniProtKB/TrEMBL entries from the UniProtKB keywords and UniProtKB subcellular locations vocabulary mappings.

Emily: this change will now occur in the March 5th UniProt-GOA release to provide groups with enough time to adapt.

Emily: the original GO_REF:0000004 and GO_REF:0000023 codes remain valid (and have been updated) and can be used by groups who would prefer to group IEA annotations originating from UniProt keywords2GO and UniProt subcellular Location2GO together.

It will just be that UniProt-GOA will be using instead GO_REF:0000037,GO_REF:0000038 instead of GO_REF:0000004 and GO_REF:0000039 and GO_REF:0000040 instead of GO_REF0000023 in releases from March onwards.

Part 2: Proposal to rename of spsl and spkw mapping files to better indicate they are UniProt vocabularies

Chris: Before making this change UniProt should make sure that the new name for these vocabularies and associated Ids are appropriately referenced in GO Consortium resources - files e.g. GO.xref_abbs and OBO libraries

Part 3: Intention to transform some incorrect IEAs supplied by the UniProt-GOA files.

Julie: this needs to be widely advertised. Perhaps include full details in a UniProt-GOA paper?

Emily: expecting that changes to IEA annotations will occur in UniProt-GOA releases a few months time.