POC Conf. Call 8-23-11
POC meeting, Webex Conference Call; Date: Tuesday Aug 23rd, 2011 10am (PDT)
In attendance:
POC members: Laurel Cooper (OSU), Ramona Walls (NYBG), Justin Elsner (OSU), Justin Preece (OSU), Pankaj Jaiswal (OSU), Dennis Stevenson (NYBG), Marie Alejandra Gandolfo (Cornell University)
Absent: Barry Smith (University at Buffalo, NY), Chris Mungall (Lawrence Berkeley National Lab)
Collaborators: none
Acceptance of the minutes from the POC_Conf._Call_8-2-11? There were no changes, additions, or deletions.
Conference report
- LC and PJ attended Plant Biology 2011, Aug 6-10th, Minneapolis, Minn
Gramene and Plant Ontology hosted a Data Curation Workshop again, focusing on Gramene tools and the PO.
LC and PJ presented a PO poster.
TAIR/PMN (Kate Dreher) organized a Plant_Biology_2011_Outreach_Boothand we took part, we hosted the website.
Update on addition of new MaizeGDB Annotations
- This was discussed on the last conference call: [POC Conf. Call 8-2-11]
- We had a webex meeting with Mary Schaeffer on Aug. 11th, see: Submission of Association files for the Kaeppler gene expression data from MaizeGDB- Aug 2011
- Mary sent us a "test" file of 25,000 lines- report from JE about loading the file re: Issues?
4015 annotations in beta browser
incorrect notation in the "assigned by" column
link to gene model page lists species as "unknown"
All of these problems seem to be due to the fact that there was a missing column in the annotation file (either column 11 or 12). We will look at the file and fix it (or have Mary fix it).
There was originally a problem with the loading script handling large files, but that has been fixed.
We may want to consider breaking up the files into smaller files, to make working with them easier. We should only load their new files when we are doing a build, since the files are so large. In theory, the computer could run out of memory while loading, although that hasn't happened yet.
May need to add new link to MaizeGDB in xref library, because now the gene just links to the general page for MaizeGDB, rather than the specific page for each gene model.
Update on getting rid of the "sensu" terms
See: Eliminating Zea/Poaceae terms from PO and the discussion from 8-2-11 meeting.
Got rid of 29 Zea/Poaceae specific terms by merging them into the their parent terms. See list of merged terms here.
Merged terms were those needed to accommodate the new MaizeGDB annotations, mostly anther and its parts, and silk/style.
This doesn't seem to cause any problems with loading the file. Still need to test how it will affect moving the annotations for the merged terms (they should automatically move with the merge).
Comments for merged terms: We should add a comment asking users to put annotation on corresponding floret type. Need to list the names and IDs for any floret types that are appropriate for the term being commented on.
Mary felt is was a good idea to add the comments, and to be sure to communicate with users about the changes.
Other users may have to alter their database structure to handle extra annotation in column 16. We need to be sure to alert our users to it.
Justin still has to replace the annotations for merged terms in the annotation files. Don't move automatically. He will need to check for annotations on alternate ID before each release. RW will send JE a list of the PO id's that have been merged. Also need to alert Mary so that she can move the annotations in their files. With each new release, JE can run a QC check to look for any lines in the association files that reference obsolete or alternate IDs.
There are still about 50 terms with Zea or Poaceae in their name that need to be merged, plus about 40 terms that don't have Zea or Poaceae in their name, such as PO:0006419 androecium of lower floret of pedicellate spikelet of ear or PO:0006433 lemma of lower floret of sessile spikelet of ear.
We agreed to merge these terms as well. Need to be sure to put comment about annotating to the corresponding floret. We will keep all of the floret types.
Also need to work on PGDSO term: 3 infloresence visible - rework the tree, add a term. Suggested changes have been detailed on SF tracker.
We will put this on the agenda for next week.
Other items that came up during the meeting
Tracking changes in PO
From DWS: As the PO file is changing, is there any way to document what we are doing and why?
It is documented in the SVN site and SourceForge, but that information is not so easy to find after some point. Better to have the information in the obo file. PJ: there are ways to document changes in obo, but it is not standard format. PJ is asking them to make modified_on and modified_by part of the standard obo format.
tassel and ear
We don't want to get rid of tassel and ear at this point (although they could just be synonyms of inflorescence). "Ear" is only used in maize. "Cob" is a synonym of inflorescence, used for sorghum.
No one was comfortable with the name "functionally male (or female) inflorescence", because it suggests that both male and female flowers are present, but only one type is functional.
We will use the names "tassle inflorescence" and "ear inflorescence".
strawberry structures
After some discussion, we decided that recepticle was the appropriate term for the fleshy red part of a strawberry and fruit (achene) is the appropriate name for the individual mature carpels (the brown things sticking out of the red part).
What is recepticle part of? Flower or inflorescence or stem? Right now it is not part_of anything, because we weren't sure where to place it. We may want to make it part of plant axis.
Note: on the PO browser, receptacle is_a inflorescence, but that appears to be a mistake. Probably it was supposed to be part_of inflorescence, but that is not appropriate for all species (some receptacles are at the base of single flowers). Need to get rid of that relation.
Eucalyptus
We have the term calyptra perianth for Eucalyptus (sometimes called an operculum). We need to add two specific children for this, one for petals and one for sepals, because this can vary by species, and some species have two calyptra. (We should also add one for tepals.)
PJ requested that we add "operculum" as a broad synonym for all of them.
Syncing PO and CL
LC will set up a conference call with Terry and Alex from CL to go over the changes described below.
CL plans to obsolete most of its plant cell types, and add xrefs to those for PO terms.
CL will provide bridge files as part of the CL release process.
Problematic xrefs to CL terms from PO
PO has xrefs to 46 CL terms.
- Most are accurate and reference CL plant cell types. Suggest that we delete these xrefs, because CL is going to obsolete those terms.
- Nine of the xrefs are not under CL's category for plant cells, and require special attention (see below).
Here is what CM said about the xrefs to CL: "Hmm, have to figure out what to do about the CL xrefs. I think all the ones that have been obsoleted should be translated to alt_ids. If we have one remaining, PO:0009002, keep the xref and add a header: treat-xrefs-as-equivalent: CL".
plant cell
PO:0009002 (plant cell) xref to CL:0000610 (plant cell).
CL defines plant cell as "A cell found in seeded plants." This is outdated.
Suggest that CL adopt PO's definition of plant cell. We should keep this xref. See below on treating xrefs as equivalent terms.
CL should probably obsolete CL:0000610 and use PO:0009002.
zygote
PO:0000423 (zygote) xref to CL:0000365 (zygote).
CL:0000365 is_a cell by lineage, is_a early embryonic cell. This is probably fine.
We should keep this xref. See below on treating xrefs as equivalent terms.
We don't need to name this "plant zygote" because it is really the same thing across organisms.
This is okay.
epidermal cells
We agreed to the suggestions below, but will set up a call with CL to go over them.
- PO:0004013 (epidermal cell) xref to CL:0000362 (epidermal cell)
CL:0000362 is_a squamose epithelial cell, is_a general ecto-epithelial cell
At an earlier meeting and via email exchanges, we decided that the CL definition of epidermal cell is not appropriate for plants.
PO:0004013 should be renamed plant epidermal cell (?). Remove xref to CL terms. PO:0004013 will be imported into CL with other plant cell types.
- PO:0000332 (epidermal pavement cell) xref to CL:0000400 (pavement cell)
CL:0000400 is_a epidermal cell. CL should obsolete this term, and PO should remove xref. PO:0000332 will be imported into CL with other plant cell types.
- PO:0000115 (socket cell) xref to CL:0000297 (socket cell)
CL:0000297 is_a epidermal cell, is_a supportive cell. CL should obsolete this term, and PO should remove xref. PO:0000115 will be imported into CL with other plant cell types.
- PO:0000282 (trichome) xref to CL:0000282 (trichome)
CL:0000282 is_a epidermal cell.
The xref from PO:0000282 to CL:0000282 should be on removed, because CL:0000282 really corresponds to PO:0008030 (trichome cell).
It is not clear from the CL definition if CL:0000282 refers to a hair in any organism or only to plant trichomes. If it refers to plant trichome cells only, then CL should obsolete CL:0000282 along with the other plant cell types.
others
We agreed to the suggestions below, but will set up a call with CL to go over them. For example, need to make sure that sheath cell really was supposed to refer to a type of plant cell and that this is not some kind of animal cell called a sheath cell. Need to talk to them about trichomes or hair cells too. Should look at how GO treats them.
note: there is a sheath cell in C. elegans
- PO:0004531 (sheath cell) xref to CL:0000618 (sheath cell).
CL:0000618 has no definition, but is_a columnar/cuboidal epithelial cell. Suggest we remove this xref and the CL obsolete the term.
- PO:0000263 (atrichoblast) xref to CL:0000332 (atrichoblast).
CL:0000332 is_a epidermal initial, is_a non-terminally differentiated cell (second is_a relation is redundant).
PO definition of atrichblast is: A root epidermal cell that is formed after asymmetric division of an epidermal initial and does not give rise to a root hair.
Suggest that CL obsolete CL:0000332, because it is a plant cell, and their definition is not up to date. We will remove xref. PO:0000263 will be imported into CL when they bridge to plant cell types.
- PO:0004011/PO:0000070 (initial cell) xref to CL:0000034 (stem cell) and CL:0000266 (meristemoid)
definition of PO:0004011/PO:0000070: A meristematic cell that by division gives rise to two cells, one of which remains meristematic, while the other is added to the plant body. Comment: Initial cells are often part of a portion of tissue that is composed mostly of somewhat older, differentiating cells. The division of an initial cell is often uneven, with the smaller daughter cell remaining meristematic.
definition of CL:0000034: A relatively undifferentiated cell that retains the ability to divide and proliferate throughout life to provide progenitor cells that can differentiate into specialized cells.
definition of CL:0000266: A cell constituting an active locus of meristematic activity in a tissue composed of somewhat older, differentiating cells.
Based on these definitions, PO:0004011/PO:0000070 is a better match to CL:0000266. Furthermore, meristemoid is an exact synonym of initial cell.
We should remove the xref to CL:0000034 and keep the xref to CL:0000266. See below on treating xrefs as equivalent terms.
PO cell types that should reference cell types in CL
A few types of cells are equivalent to remaining CL terms. If so, then they should have a general xref, then assert the general xrefs as equivalent using the header "treat-xrefs-as-equivalent: CL" (see obo format guidelines).
PO:0009002 plant cell > CL:0000610 plant cell
PO:0000423 zygote > CL:0000365 zygote
PO:0004011/PO:0000070 intial cell/meristemoid > CL:0000266 meristemoid
This is okay for zygote and meristemoid. Maybe CL should get rid of CL:0000610.
New terms for CL
There are four terms in CL for which PO does not have equivalent terms, because it was set up for only angiosperms. Suggest we add terms like these when we work on the revisions of conducting tissue and cells. For now, Terry has removed the problematic xrefs from CL, will update them if we add new terms.
See conducting cells for non-vascular plants.
New subset for CL
RW has created a new subset for plant cells. It contains all of the is_a descendents of plant cell. CL can use this to create their bridge file.
This subset will need to be updated whenever we add new cell types.
OK
Upcoming meetings 2011:
- POC Meeting at New York Botanic Garden Dates, Sept 10th-11th, 2011
The PO meeting will be on Saturday Sept. 10 and Sunday Sept. 11th.
RW has reserved the NYBG apartment for the nights of Sept. 9 and 10th, but it looks like the Hotel in Yonkers will work better for LC, PJ and BS -Reservation for the Garden apartment has been cancelled.
We are working on developing the agenda: See: POC Meeting at NYBG; Sept 10th-11th, 2011
