[ome-users] CZI channel description

[Roger Leigh]

On 12/07/14 10:04, Emil Rozbicki wrote:
> Forwarded response from Forrest Collman:
>
> Hi Emil,
>
> thank you so much for taking a look at this issue, and making this
> available via the original metadata to me. I think i can use this to
> make my workflow work correctly.
>
> However, I respectively disagree about the channel names in Zen.  It
> seems to me that the schema should be able to specify what the meaning
> of the channel that is acquired is, and that this information is
> critical to understanding what the image contains.  So with respect to
> Alexa 594 vs ASIC1... Alexa 594 corresponds to a fluorophore that Zen
> has preprogramed.. which corresponds to a choice of filter cube
> optimized to observe that fluorophore.  The user cannot enter ASIC1 as
> the channel name in ZEN, as ASIC1 is biological molecule, and ZEN
> doesn't know apriori what filter cube to use with it.  Alexa 594 is the
> fluorophore is attached to ASIC1 molecule of interest.  This is the
> meaning of the biological image.  To not have a way to include this in
> the OME-XML model seems fundamentally flawed.

I have opened a ticket for this: please see
https://trac.openmicroscopy.org.uk/ome/ticket/12473

However, I should note that while the OME-XML data model doesn't
currently support it, this is primarily due to the fact that it is
modelling what the majority of existing file formats provide.  Most
formats only provide a single name for a channel, and CZI differs here
in providing additional metadata which doesn't fit.  This isn't
fundamentally flawed; it's a case where the model requires extending to
support the richer metadata provided by newer file format.

> Put another way, our microscope only has 4 filter wheels.  Every image
> it takes will have those channel names in ZEN, but we take images using
> hundreds of different antibodies which are tagged with different
> fluorophores. If the OME-XML schema doesn't let me write down what
> antibodies i was using when i snapped a picture, how it a useful tool to
> keep track of my data.
>
> In ZEN, its only this description field that the user can use to input a
> "meaningful" name to display, that keeps track of what the true
> biological meaning of the image is.  That seems to me, to fit the
> description of the schema, much better than the name of the
> dye/filterset that was used in the image. Perhaps you still disagree,
> but there should be somewhere else that it exists within the data?

There is no disagreement that this data is useful and needs to be
stored.  The data model needs to be general enough to cover most common
uses and file formats.  In this specific case, you are using the
description to encode the probe, but this isn't actually mandated by
either the acquisition software or the file format, so can't be reliably
modelled as such.  I've put some of the details here onto the ticket
linked above; if you have any thoughts or suggestions for this, we would
be happy to hear them.


Regards,
Roger Leigh

--
Dr Roger Leigh -- Open Microscopy Environment
Wellcome Trust Centre for Gene Regulation and Expression,
College of Life Sciences, University of Dundee, Dow Street,
Dundee DD1 5EH Scotland UK   Tel: (01382) 386364

The University of Dundee is a registered Scottish Charity, No: SC015096