Indication of cells that have been perturbed

[adirsolo]

We have a simple question regarding the Cell table in the sql database of the extracted features: we are trying to understand which cell has been perturbed and which not. We observed that this information (e.g., "treated" or "mock") was found with the well resolution (in the table of "mean_well_profiles") but we have not found any indicator with the cell resolution. We also tried to map the cells to their corresponding wells but we have not managed to detect any common key field or something like that.
We are probably missing something and we hope that you can help us to figure out what.
Thanks!

[shntnu]

Are you able to map each cell to the well from which it comes? If not, see this that shows you how to do it.

Once you have the cell-to-well mapping, you can then use the Metadata_* columns from profiles.tar.gz (see #3 (comment)) to get the well-to-compound mapping (there may be an easier way but this works).

[shntnu]

I was emailed this question:

Q. My team is interested in using the Bray 2017 et. al. sample Cell Painting data available on GitHub; however, after downloading it we found that based on the naming scheme used, it isn’t obvious which plate corresponds to what condition—we must be missing something here.

A. See the second part of my comment #9 (comment) to get the well-to-compound mapping

[shntnu]

Another related question

Q. how can one map the compounds to the images provided by the download script (download_cil_images.sh)?
For example, I see that the plate 24279, contains information about many compounds, including e.g. ajmalicine. Is there a way that I can know which images correspond to a given perturbation? I see that there is a well annotation in the image filenames, but it is not clear to me how to associate that with the metadata in the file chemical_annotations.csv. 

A. See the second part of my comment #9 (comment) to get the well-to-compound mapping