I’m working with a dense culture of neural cells where a small proportion of them express fluorescent protein GFP. In each image, I have > 2500 neurons where only ~10 are fluorescent which should help to analyse the morphology of the cells. r01c01f01p01-ch1sk1fk1fl1.tiff (1.9 MB) r01c01f01p01-ch2sk1fk1fl1.tiff (1.8 MB)
I have generated a pipeline that is able to segment the nuclei, neuronal soma and neurites using CellProfiler 4.0.7 and I use the module MeasureObjectSkeleton in order to measure the neurite branching. On top of that, I measure the morphological properties of the cells using the modules measure objectsizeshape, granularity and intensity and finally export all the data to a SQL database to classify the different objects by their morphology using CellProfiler Analyst.neurite_tracing.cppipe (27.5 KB)
I’m finding the following difficulties:
Because the culture is very dense, I can’t manage segment the neurons directly from the nucleus. I tried using the soma of the cells as a mask to reduce the number of nuclei in the field, but I still have 3-4 nuclei in the mask because the cells are very packed. I’m not sure if this is an issue or not since I can still use the XY position of the cell soma as the primary object.
In the module ExportToDatabase, I cannot select the option to export all the measurements to the database. Unless the option selected is “none” the module gives an error.
I am unable to launch CellProfiler Analyst with the output of this pipeline. When I try to load the .properties file, CPA throws an error. test_neurites.properties (6.9 KB)
I would really appreciate any help, comments and criticism on this pipeline
Thank you everyone in advance!