eric-czech
commented
5 years ago Hey @benjaminyellen (cc: @jmotschman),
I added some features to process fluorescent images, and I was going to try to hold off on pushing to master until https://github.com/hammerlab/celldom/issues/71 is done (which would automatically run tests on commits to make sure changes like this don't mess with other experiments).
All the same, you can see some results (and the commands) from an analysis on the multi-color dataset you shared here:
MLadder Multi-Color Dataset Analysis MLadder Multi-Color Dataset Commands
Let me know what is most interesting in these types of experiments, but at a minimum that's at least a look at things like the CY5 vs GFP distributions for all cells at time 0:
And what apartments look like with one of each cell type (cells detected as GFP+ shown with green circle and CY5+ with red circle):
There are also some tables in the notebook with apartment counts (and percentages) for counts of each cell type (i.e. to answer the question "what is the frequency of apartments w/ 1 GFP+ cell and 1 CY5+ cell").
What are some other calculations/visualizations you think might be common to most if not all experiments using extra fluorescent channels like this?
yellenlab
Per https://github.com/hammerlab/celldom/issues/43, fluorescent images would be separate files taken at different (nearby) times. Most likely, it will make sense to modify Acquisition with named intensity images to then be used for integrated/mean values across brightfield cell masks.