Menu
▾
▴
Re: [Obo-anatomy] [Obo-cell-type] Re: neural crest cells
|
From: Melissa H. <mha...@uo...> - 2006-06-14 18:47:57
|
Hi, I have another topic for discussion for all you cell and anatomy folks out there. We have an increasing need for adding cell types with specific names to our database. Would these be anatomical terms in ZFIN's AO, or would they go into the cell ontology? How will the cell ontology deal with organisms with identified cells (cells that can be specifically identified in each organism). For example, we would have the CaP motoneuron in somite 3. It is_a primary motoneuron, but we'd like to be able to annotate to it specifically. I know that flybase will have a similar problem. Is fly using the cell ontology yet? Any more insight on the issue of how to incorporate cell types into anatomy would be greatly appreciated. No one has commented much on the neural crest emails below, so I would like very much to hear how the other MODs are dealing with this issue. Thanks, Melissa Melissa Haendel wrote: > Chris, I am wondering if what we are doing by adding cell types to our > AO is the same as what you show in the type references section of your > pheno-syntax document. (Which, btw, is a major accomplishment!) These > mappings wouldn't be part of either ontology, but rather are a set of > Es unto themselves. I guess the question is how to handle them in > searches down the road. > > m > > Melissa Haendel wrote: >> Ok, I think we are actually on the same page. I've interspersed my >> comments below. >> >> chris mungall wrote: >>> >>> On May 25, 2006, at 1:56 PM, Melissa Haendel wrote: >>> >>>> Hi Minna, Oliver, >>>> I wanted to make a request for two new cell types in the CL, but >>>> thought my reasoning required some additional input so I am >>>> emailing both the anatomy and the cell listserves. Please feel >>>> free to forward to anyone else you feel might have a vested >>>> interest.... >>>> >>>> At ZFIN, we are currently adding cell types from the CL to our >>>> anatomical ontology (AO). We retain the CL IDs so that we will be >>>> able to support the CL independently in the future. We will keep >>>> the mapping between the CL and our AO, which is the value added >>>> beyond just using the CL. >>> >>> Hi Melissa >>> >>> There is a slight problem here. You will end up with links such as: >>> >>> spermatogonium part_of zebrafish testis >>> >>> because CL is species-neutral and your AO is species (or at least >>> fish) specific >> >> yes, but because we would retain this mapping here at ZFIN, couldn't >> we just do it this way for now, and then when we can support a >> generic DAG, keep this mapping as a separate file? then it would be >> implied that all CL terms we were using would necessarily be applied >> to the zebrafish in our mapping file. >>> >>> >>> There will be no consequences when used purely in a zebrafish >>> context, but in a big integrated database there will be problems >>> >>> We would want either >>> >>> fish spermatogonium part_of fish testis >>> >>> or >>> >>> spermatogonium part_of testis >>> >>> The latter approach will require a multi-species reference AO down >>> the testis level which will take some time. So the first approach >>> would be better for now. We can come up with some way of doing this >>> automatically so that you don't end up doing redundant work with CL, >>> so carry on with your approach of importing CL IDs, but it's >>> important to bear in mind for the future >>> >>>> The two cell types are "pre-migratory neural crest cell" and >>>> "migratory neural crest cell." The reasoning is as follows: >>>> We currently have neural crest cell as part_of the neural crest. >>> >>> Wouldn't it instead be: >>> >>> pre-migratory neural crest cell part_of neural crest >> exactly why we need this request from the cell ontology. The >> migratory neural crest cell would NOT be part_of the neural crest. >>>> Therefore, searches for genes expressed in the neural crest would >>>> return annotations to neural crest cell(s) outside of the neural >>>> crest itself. >>> >>> which is a problem, and would be fixed by the above >>> >>>> We also have two types of neural crest terms in our AO right now, >>>> for example, trunk neural crest, which refers to a positional >>>> subset of the neural crest, and facio-acoustic neural crest, which >>>> refers to a migrating population of cells. The latter is not >>>> really part of the neural crest, but rather part of the head (or >>>> whatever) and is_a migrating neural crest cell. >>> >>> this seems wrong - you can't have a population of cells stand in an >>> is_a relation to a type representing a single cell. Why not have >>> >>> facio-acoustic NC cell is_a migratory cell >>> facio-acoustic NC cell part_of facio-acoustic NC >>> >>>> >>>> The confusion here is that often neural crest is used to refer to >>>> both a place and sets of migrating cells, which are really present >>>> in different anatomical locations. >>> >>> well you can't have one term referring to two distinct types. Either >>> you have to come down on one side, or create two new terms, and >>> demote "neural crest cell" to a synonym of each. The terminological >>> confusion is definitely an issue, but it can be separated from the >>> ontology. >> >> Yes, one needs to use facio-acoustic NC cell instead of >> facio-acoustic NC, which doesn't really exist because it refers to a >> location outside of the neural crest. I will make changes to our AO >> such that we only have neural crest terms that refer to a part of the >> neural crest. the migrating neural crest cells should really be >> cross products with other body parts like branchial arch 1, or be >> subtypes that traverse a specific pathway, like facio-acoustic neural >> crest cell (we would need new CL terms in this case) OR maybe the >> solution is really to have a term in our AO like facio-acoustic >> neural crest cell population, because we are not really talking about >> a cell type or a tissue- its really in between. The bigger question >> is how to describe migrating populations of cells. >> >> Thanks for the instant brainstorming! >> melissa >>> >>>> Anyone else had this type of problem? how are the other MODs >>>> dealing with incorporating cell types into their anatomical searches? >>>> >>>> Thanks for the brainstorming. >>> >>> cheers >>> c >>> >>>> Melissa >>>> >>>> -- >>>> Melissa Haendel, Ph.D. >>>> >>>> >>>> ZFIN Scientific Curator >>>> >>>> Zebrafish Information Network >>>> >>>> 5291 University of Oregon >>>> >>>> Eugene, OR 97403-5291 >>>> >>>> Phone: (541) 346-5108 >>>> >>>> >>>> >>>> ------------------------------------------------------- >>>> All the advantages of Linux Managed Hosting--Without the Cost and >>>> Risk! >>>> Fully trained technicians. The highest number of Red Hat >>>> certifications in >>>> the hosting industry. Fanatical Support. Click to learn more >>>> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=107521&bid=248729&dat=121642 >>>> >>>> _______________________________________________ >>>> Obo-anatomy mailing list >>>> Obo...@li... >>>> https://lists.sourceforge.net/lists/listinfo/obo-anatomy >>> >>> >>> >>> ------------------------------------------------------- >>> All the advantages of Linux Managed Hosting--Without the Cost and Risk! >>> Fully trained technicians. The highest number of Red Hat >>> certifications in >>> the hosting industry. Fanatical Support. Click to learn more >>> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=107521&bid=248729&dat=121642 >>> >>> _______________________________________________ >>> Obo-cell-type mailing list >>> Obo...@li... >>> https://lists.sourceforge.net/lists/listinfo/obo-cell-type >> > -- Melissa Haendel, Ph.D. ZFIN Scientific Curator Zebrafish Information Network 5291 University of Oregon Eugene, OR 97403-5291 Phone: (541) 346-5108 |