#436 gene specific db x_refs

[Valerie Wood]

we would sometimes like to make gene specific links from the database cross references section

We will capture these, for now in artemis qualifier

Syntax

/controlled_curation="term=external link; type=
where
type = the description of the link which will appear on the gene page
the accession will usually be of the form

GEO:GSE57040
ArrayExpress:E-MTAB-2543

To do,
1. check that these are the correct prefixes in the DBxrefs file

1. open corresponding jira ticket

2. Add curator documentation
   (this should specify a list of types
   so far we have
   deletion mutant expression profiling
   WT overexpression expression profiling
   genome-wide localization

3. Migrage existing
   misc, deletion mutant expression profiling

4. Let users know that they can submit gene specific db xrefs to their array experiments
   (we could have a specific submit form where they can supply gene name, type, accession

5. Wonder why we can't automate this via a query on arrayExpress or GEO

[Valerie Wood]

/controlled_curation="term=external_link, description; db_xref=GEO:GSE57040"

at present the cv "external_link" will only have 2 terms:
deletion mutant expression profiling
WT overexpression expression profiling

the term will be the description in the external refs section of the gene page

This will be extendible as potentially we can use with other DBs and different descriptions.

[Antonia Lock]

I also added genome-wide localization to your list above Val.

This is what I entered for the nts1 (SPCC24B10.19c) record

/controlled_curation="term=external_link, WT overexpression expression profiling; db_xref=ArrayExpress:E-MTAB-2543"
/controlled_curation="term=external_link, genome-wide localization; db_xref=GEO:GSE57040"

also added to nse6 because it is from the same paper:
/controlled_curation="term=external_link, deletion mutant expression profiling; db_xref=ArrayExpress:E-MTAB-2543"

In the comment Nicola said that that dataset should also cover deletion mutant expression profiling but that is not in the file descriptions....I'll doublecheck that one.
Note for future self
/controlled_curation="term=external_link, deletion mutant expression profiling; db_xref=ArrayExpress:E-MTAB-2543"

[Valerie Wood]

OK great.

I guess we need to keep this "cv" of "descriptions" in a file somewhere (we are treating it like ontology terms to make the chado loading easier).
Can somebody make that happen, and tell Kim name/location (embarrassingly I don't know how)

[Kim Rutherford]

Currently only GO and phenotype terms are checked against ontologies when reading the contig files. For everything else the term (cvterms in Chado) are created as needed. You can browse them here:
http://curation.pombase.org/pombe/view/list/cv?numrows=500&page=1&model=chado

Here's an example:
http://curation.pombase.org/pombe/view/object/cv/46?model=chado

This is because when we started loading the contig files the they were the only source for most of the cv terms.

I've added "external_link" as an allowed terms will be magically added and you'll be able to check the terms using the link above after the next load.

I'm running a load now.

Thanks Antonia for adding some example external_links. They are appearing here:
http://curation.pombase.org/dumps/builds/pombase-build-2015-02-09-v3-l1/logs/log.2015-02-11-09-58-11.protein_family_term_annotation

but they should disappear from the next log.

(That's the first time I've looked at that file. Sorry it's so unreadable. I'll tidy it a bit)

https://github.com/pombase/pombase-legacy/commit/ec0546fc46eb25adb13b52e2eec2356e2eeeeaee

[Kim Rutherford]

Antonia's test external_links are in the latest load:
http://curation.pombase.org/pombe/view/object/cv/external_link?model=chado

[Antonia Lock]

new jira
https://www.ebi.ac.uk/panda/jira/browse/PB-2215

[Valerie Wood]

Follow on for curators:
https://sourceforge.net/p/pombase/curation-tasks/563/
https://sourceforge.net/p/pombase/curation-tasks/564/