InCell file names

Does anyone know how to load images from a GE InCell analyzer into Cell Profiler using metadata? Here are some examples of file names. Row B, Column 2 or 3, channel Blue, Green and DIC.

B - 2( wv Blue - FITC).tif
B - 2( wv Green - dsRed).tif
B - 2( wv TL-DIC - dsRed).tif
B - 3( wv Blue - FITC).tif
B - 3( wv Green - dsRed).tif
B - 3( wv TL-DIC - dsRed).tif

As it turns out, I was working with InCell files just last week… :smiley:

The regular expression I used (assuming that’s what you’re referring to) is this:^(?P<Row>.*) - (?P<Column>[0-9])\( wv (?P<ChannelColor>^\s]*) - (?P<ChannelFilter>^\s]*)\).tif

EDIT: Oops - Choose which answer you like, Mark’s or mine! We posted almost at the same time.

The simplest way is to extract the metadata from the file headers: In Metadata, under 'Metadata extraction method" choose “Extract from image file headers”. However the metadata may be wrong or incomplete.
So try this:

The difficult bit is that the parentheses characters need to be escaped.

Hope that helps,

Thanks very much Mark and David. I will try this out later today and report back. My short term solution was to do a batch file name change to get rid of the parenthesis and change the channel names to numbers. I think a long term solution with the InCell and Cell Profiler would be to convince GE to change their file name conventions? Any thoughts from InCell users, or GE?

Unfortunately;y, it’s probably pretty unlikely that you’ll get a big corporation to budge on what they’ll consider a minor thing… :smile:

We are in the process of adding a feature to the regular expression tool in the Metadata module to try to guess the appropriate expression. That involves curating the known screening microscope filename formats, so it may take some time to complete.


I think it’s unlikely as they sell their own cell analysis software. They wouldn’t want to cut into profits by making it too easy to use competing software!

The latest versions (v7.2) of the IN Cell acquisition software have an option for the file name. There’s the old format as above, the new default that’s almost identical but pads fields with zeros to help sorting, a parentheses free version of the same, which should be easier to parse since it loses special characters, and a new version that lists all tokens (field, z, time, etc) no matter what: A_01_f2_c1_xUV_mDAPI_z3_t2.tif


1 Like