Follow up meeting: Sept 15, 2011
POC members: Laurel Cooper (OSU), Ramona Walls (NYBG)
Collaborators: Mary Schaeffer, MaizeGDB
Items for discussion
Formatting of Association Files
- Question from MS @formatting column 16 correctly in last send - does it matter? You list on the po how to page an example: part_of(PO:0009025)
The formatting for column 16 is correct
- Adding the duplicate annotations for items also in column 16 will be done. Copy all the lines and replace the primary PO:ID with the one referred to in column 16.
This will affect the following terms:
- PO:0009066 anther: part_of(PO:0006310) (tassel floret; for anthers R1 B73 )
- PO:0009054 inflorescence bract: part_of(PO:0020136) (ear inflorescence; for husk, innermost R1 B73, husk, innermost R2 B73, and husk, outer R2 B73)
- PO:0006339 juvenile leaf: part_of(PO:0009025) (vascular leaf; for leaf 1st with sheath V3 B73)
- PO:0020040 leaf base: part_of(PO:0009025) (vascular leaf; for leaf base of expanding leaf V7 B7)
- PO:0006340 adult leaf: part_of(PO:0009025) (vascular leaf; for leaf, 13th V9 B73)
- PO:0009074 style: part_of(PO:0006354, or use PO:0006350, PO:0006351, PO:0006352 or PO:0006353 if appropriate) (for silks R1 B73)
(PO:0006354=ear floret; PO:0006350 =upper floret of pedicellate spikelet of ear; PO:0006351=lower floret of sessile spikelet of ear; PO:0006352=upper floret of sessile spikelet of ear; PO:0006353=lower floret of pedicellate spikelet of ear)
- A question about NULL values, may I use the term 'NULL'? Or do you require that there be a blank? These should be a hyphen
The MaizeGDB files are built in their OracleDB then moved to Access, then to Excel. So it may have been an issue about crossing the software platforms. Mary will convert the Nulls to hyphens from the Oracle Tables. Their Databases are updated monthly. MS problem finding errors in perl.
After the call: JE sent MS a command line script to remove the commas and replace with tabs
Should we add this on the Documentation Page as a general guideline?
- Request from MS: A tab-delimited file of all the 'obsolete' PO terms
she does not need this anymore as the obsoleted terms now have 'obsolete' in the definition.
- Test file on the SVN: po_anatomy_gene_zea_MaizeGDB_silk.assoc.txt
MS will download it and check it and add the additional lines needed to refer to "part_of(PO:0006354), which is ear floret
file will be saved as: po_anatomy_gene_zea_MaizeGDB_silk.assoc - to be consistent, don't need the .txt
Once all the bugs are worked out of this file, the rest of the annotations will be submitted in two files:
po_anatomy_genemodel__zea_MaizeGDB.assoc (use aspect A)
po_development_genemodel__zea_MaizeGDB.assoc (use aspect D)
Links to the Zm GeneModels and tissue pages
They will have same URL structure/link unless the models go away or change, and then it will be by synonym or something. They seem not to be changing too much between versions from the early reports - mostly coordinates.
From MS: What would be the best way to provide the maize tissue links? Can we use column 8 for example and supply in this form: MGDB_tissue:2366347, The URL would be the standard one for MaizeGDB – just by ID
Another way: In the reference field, column 6, add the tissue ID (which links to the reference in the above link) - this is logically more consistent with GO notions I would think.
Neither of these places in the annotation file will work. LC will create a page on our wiki with links to the tissue pages, and to a description of the experiments
Need to figure out why the PubmedID is not showing up
- Display of the Pubmed Publications:
MS put in both the pubmed ID and the maizeGDB ID, as recommended, but only the maizeGDBID is showing up, but it is not creating a link there (check dbxrefs) and it is not displaying or linking to the pubmed citation. See what is up with this? If the maizeGDBID link was working it would link to the tissue pages from there
- Linking to the gene model pages:
The link in the "Assigned by" column should take you to the Gene model page. This is determined by the Dbxrefs file. Need to add an entry for the maize gene model name. MS will check with her person to find out how to link to these (RW: see bottom link on the list below. This is the current link to the gene model pages. MS should confirm that this is stable.)
Dbxrefs file:
current contents
PO has two stanzas for MaizeGDB in the dbxref file:
abbreviation: MaizeGDB
database: Maize Genetics and Genome Database.
object: Canonical name of gene.
example: MaizeGDB:apb1
generic_url: http://www.maizegdb.org/
url_syntax: http://www.maizegdb.org/cgi-bin/displaylocusresults.cgi?term=
object: Phenotype of gene variant
example: MaizeGDB:liguleless
url_syntax: http://www.maizegdb.org/cgi-bin/displayphenoresults.cgi?term=
abbreviation: MaizeGDBID
database: Maize Genome Database.
object: locus Identifier.
example: MaizeGDBID:12008
generic_url: http://www.maizegdb.org/
url_syntax: http://www.maizegdb.org/cgi-bin/displaylocusrecord.cgi?id=
object: phenotype identifier
example:MaizeGDBID:61504
url_syntax:http://www.maizegdb.org/cgi-bin/displayphenorecord.cgi?id=
Note that both of these have a mistakes -- they contain multiple url syntaxes (is this allowed?) and do not provide examples.
proposed contents
Need to consult with Mary about what should link where.
Under construction
"canonical name of gene" and "Phenotype of gene variant":
abbreviation: MaizeGDB
database: Maize Genetics and Genome Database.
object: Canonical name of gene.
example: MaizeGDB:apb1
generic_url: http://www.maizegdb.org/
url_syntax: http://www.maizegdb.org/cgi-bin/displaylocusresults.cgi?term=
object: Phenotype of gene variant
example: MaizeGDB:liguleless
url_syntax: http://www.maizegdb.org/cgi-bin/displayphenoresults.cgi?term=
"locus identifier" and "phenotype identifier":
abbreviation: MaizeGDBID
database: Maize Genome Database.
object: locus Identifier.
example: MaizeGDBID:12008
generic_url: http://www.maizegdb.org/
url_syntax: http://www.maizegdb.org/cgi-bin/displaylocusrecord.cgi?id=
object: phenotype identifier
example:MaizeGDBID:61504
url_syntax:http://www.maizegdb.org/cgi-bin/displayphenorecord.cgi?id=
gene model pages:
abbreviation: MaizeGDBID_genmod
database: Maize Genome Database
object: gene model name
example: MaizeGDBID_genmod:GRMZM2G172795
generic_url: http://www.maizegdb.org/
url_syntax: http://www.maizegdb.org/cgi-bin/displaygenemodelrecord.cgi?id=
url_example: http://http://www.maizegdb.org/cgi-bin/displaygenemodelrecord.cgi?id=GRMZM2G172795
tissue pages (if needed):
abbreviation: MaizeGDBID_termref
database: Maize Genome Database
object: tissue description
example: MaizeGDBID_termref:2366347
generic_url: http://www.maizegdb.org/
url_syntax: http://www.maizegdb.org/cgi-bin/termrefs.cgi?id=
url_example: http://www.maizegdb.org/cgi-bin/termrefs.cgi?id=2366347
Maize tissue Pages:
MGDB_tissue:2366347 anthers R1 B73
MS will update the links on the tissue pages that refer to the PO terms and will link back to that page from the associated linked page: [1]
LC will create a page on our wiki with links to the tissue pages, and to a description of the experiments
Merging all of the Zea/Poaceae terms into their parent terms
Detailed all of the changes on our wiki page at: Eliminating_Zea/Poaceae_terms_from_PO
MS has already used the more general terms for these.
- Updating the existing association files: MS will look after updating these files. LC will let JE know. done
for:
- new annotations
- merged terms
- obsoleted terms
These changes need to be made in the maizeGDB database as well, so they won't have to be updated every time MaizeGDB updates their files. In each case, the lines of the annotation file should be changed to the new PO term.
List of merged terms
Note that in some cases, there are two layers of merging, so, for example, the annotations for PO:0006472 Zea stamen and PO:0006441 Poaceae stamen both need to be transferred to PO:0009029 stamen.
term (# annotations) > term merged into (= target term)
PO:0006337 inflorescence bract of ear (26) > PO:0009054 inflorescence bract
PO:0006472 Zea stamen (1) > PO:0006441 Poaceae stamen (7) > PO:0009029 stamen
PO:0006473 Zea anther (363) > PO:0006442 Poaceae anther (6) > PO:0009066 anther
PO:0006487 Zea stigma (56) > PO:0006468 Poaceae stigma (0) > PO:0009073 stigma
PO:0006488 silk (56) > PO:0006469 Poaceae style (100) > PO:0009074 style
PO:0006445 Poaceae anther wall (2) > PO:0000002 anther wall
PO:0006455 Poaceae carpel (3) > PO:0009030 carpel
PO:0006446 Poaceae endothecium (1) > PO:0020002 anther wall endothecium
PO:0006318 Poaceae floret (71) > PO:0009082 floret
PO:0006329 Poaceae floret meristem (1) > flower meristem PO:0000229
PO:0006384 Poaceae gynoecium (3) > PO:0009062 gynoecium
PO:0006465 Poaceae integument (3) > PO:0020021 integument
PO:0006508 Poaceae integument epidermis (1) > PO:0006043 integument epidermis
PO:0006464 Poaceae megaspore (1) > PO:0020019 megaspore
PO:0006463 Poaceae megasporocyte (2) > PO:0000431 megasporocyte
PO:0006444 Poaceae microsporangium (2) > PO:0025202 microsporangium
PO:0006452 Poaceae microspore (22) > PO:0020048 microspore
PO:0006451 Poaceae microsporocyte (1) > PO:0020047 microsporocyte
PO:0006462 Poaceae nucellus (2) > PO:0020020 nucellus
PO:0006456 Poaceae ovary (1) > PO:0009072 ovary
PO:0006457 Poaceae ovule (5) > PO:0020003 ovule
PO:0006497 Poaceae stigma epidermis (3) > PO:0006061 stigma epidermis
PO:0006449 Poaceae stomium (1) > PO:0020101 stomium
PO:0006450 Poaceae tapetum (3) > PO:0009071 anther wall tapetum
Open Source Forge tracker items
MS: Could you please look at the following tracker items, and post a comment on them if you think there are any problems:
See also the PO responses on the tracker items for:
IL.03 full inflorescence length reached
These are all ok, MS is fine with what we decided to do with them
MS says that the PO term radicle emergence is more applicable, so she will use that term, rather than coleoptile emergence. BTW: in corn seed the radicle emerges from the pericarp, rather than the seed coat.
