[ome-devel] WellSample LSID (Cellomics )

[Josh Moore]

On Fri, Feb 24, 2017 at 5:23 PM, Jäälinoja Harri
<harri.jaalinoja at helsinki.fi> wrote:
> Hi Josh,

Hi Harri,


> On 24/02/17 11:45, Josh Moore wrote:
...

> I just imported a plate like this:
>
> hajaalin at dhcp-asv-66:/mnt/lmu-archive/LMU-archive2/users/FROM_CELLINSIGHT/LMU-CELLINSIGHT_140628160001$
> ls *A01*| sed -e 's/f/f /' -e 's/d/ d/'|sort -n -t' ' -k2 |head -2
> LMU-CELLINSIGHT_140628160001_A01f 67 d0.C01
> LMU-CELLINSIGHT_140628160001_A01f 67 d1.C01
> hajaalin at dhcp-asv-66:/mnt/lmu-archive/LMU-archive2/users/FROM_CELLINSIGHT/LMU-CELLINSIGHT_140628160001$
> ls *A01*| sed -e 's/f/f /' -e 's/d/ d/'|sort -n -t' ' -k2 |tail -2
> LMU-CELLINSIGHT_140628160001_A01f 126 d0.C01
> LMU-CELLINSIGHT_140628160001_A01f 126 d1.C01
>
> So there should be a well_index of 126, but the maximum is 59. I understood
> from an example Ola sent earlier that the Hibernate object wouldn't have it,
> but it's strange that it's not in the db either. The CellomicsReader code
> looked to me like it would store the original id.

Ah, sorry. I misunderstood what you were looking to do. The field
index value from the Bio-Formats worlds (as opposed to what it becomes
in OMERO) is only stored in the Image name: "Well %s%02d, Field #%02d"
(see [1]). So for the moment, you would have to parse it from there.

Cheers,
~Josh.


[1] https://github.com/openmicroscopy/bioformats/blob/c0bfbe2c7c7a3e1dc6de1cb99e246068cef9042c/components/formats-gpl/src/loci/formats/in/CellomicsReader.java#L363

>>> The point is that fieldIndex is read from the filename, and may be
>>> different than fieldCntr. For example if the fields are collected in a
>>> spiral pattern that does not start in the center, the first fieldIndex may
>>> be e.g. 50. OMERO plate viewer shows fieldCntr (or some index starting at
>>> 1). If I want to find individual original files that contributed to a field,
>>> I need to find this mapping from fieldCntr to the original field index.
>>
>> You can also attempt to use omero.cmd.UsedFilesRequest:
>>
>>
>> https://github.com/openmicroscopy/openmicroscopy/blob/develop/components/blitz/resources/omero/cmd/FS.ice#L85
>>
>> which will give you a list of files that match the particular series.
>> You can see an example of its usage here:
>>
>>
>> https://github.com/mtbc/omero-downloader/blob/dev_5_3/src/main/java/org/openmicroscopy/client/downloader/Download.java#L239
>
> I didn't try this yet, but I suspect that if UsedFiles is same as
> Fileset.usedFiles, the problem will be that when a screen is uploaded, each
> of the displayed images refers to the entire set, not only the channels of a
> particular field.
>
> It's tempting to try to find the filenames, because I'd be able to process
> the data without transferring them. But then again, these are not huge
> datasets, so probably transferring is not a problem. I'll try that approach
> (CellProfiler data lists with omero iids) next.
>
> Thanks, and have a good weekend!
> Harri
>
>>
>>> Thanks, best regards,
>>> Harri
>>
>> All the best,
>> ~Josh