[ome-users] Re: Developmental stage

[Ilya Goldberg]

On Jul 19, 2005, at 4:25 PM, Graham Klyne wrote:

> Ilya,
>
> thanks for your other comments, which all seem sensible and helpful to 
> me at this stage.  I'm responding to just one point here...
>
> At 12:35 19/07/05 -0400, Ilya Goldberg wrote:
>>> - developmental stages showing expression to be drawn from a fixed
>>> vocabulary (based on GO).  Each possible stage to be a CategoryGroup,
>>> each with Categories corresponding to Yes/No.  Thus any combination 
>>> of
>>> development stages can be captured and used as basis for grouping.
>>
>> A given image would not be associated with more than one development
>> stage, would it?
>
> In our case (as I understand the biology) it can.  The in situ images 
> are of whole testes, with different developmental stages of 
> spermatogenesis appearing at different places within the testis 
> structure.   Thus a single image might, in principle, show that a gene 
> is expressed during several different stages of sperm development.

Ah, well.  That means that what you really want is to categorize 
regions of the image.
We have these sub-image things called Features, which are part of the 
core hierarchy.  Features are almost completely abstract, but they can 
be assigned attributes like images and datasets.  You could make a new 
ST called FeatureClassification, and give this feature some attributes 
to describe where it is (using Location, Extent, ROI, etc).  There are 
some built-in facilities to deal with Features - for example, if you 
use the Location attribute, the SVG viewer can generate graphic 
overlays of where they are located (all thanks to ST re-use).

If you want to assign an image to multiple categories without 
describing that these categories pertain to different parts of the 
image, that's certainly possible (though maybe sub-optimal) - you just 
can't do it with categorization.  You can make a very similar set of 
STs that don't carry this mutual exclusivity restriction.  The 
category/category group/classification are all three just regular STs 
with a fairly small API to enforce the mutual exclusivity.  If you 
don't want that, then you can easily make your own STs without this 
API.

I wouldn't get too hung up on categories/category groups.  Nobody said 
they were going to solve everyone's problems.  For this we have the 
more general STs which are in fact designed to solve everyone's 
problems ;)


-I


>
> #g
>
>
> ---
> Graham Klyne
> Image Bioinformatics Research Group (http://www.bioimage.org/)
> Department of Zoology, University of Oxford
> South Parks Road, Oxford OX1 3PS, UK
> E-mail: <Graham.Klyne at zoo.ox.ac.uk>
> Direct phone: +44-(0)1865-281991
> Departmental fax: +44-(0)1865-310447
>