Phospho akt and other membrane protein

I can’t measure phospho akt and other membrane protein with CellProfile. The propagation function doesn’t work well to separate two cells, and the N-distance can’t really include the target protein, also, both way yield less cell number compare with the nucleus, I think it is because CellProfiler automatically delete the overlapped cytoplasm, consider two very close cells as one. Looking forward to having a upgrade pipeline to solve this issue. :smiley:



Hi,

How does my attached pipeline look to you? The only odd thing about this image set was that the pAKT image has a very few bright pixels. So it seems dark unless you choose Image Contrast = Log Normalized, and thus in IdentifySecondaryObjects I chose 3-class Otsu. This can help take into account a class of pixels that are much brighter than the majority of the foreground, as well as including the main foreground pAKT class.

Hope that helps,
David
DL_pAKT_pipe.cppipe (7.22 KB)

P.S. It’s hard for me to tell by eye the separation between cytoplasm, so IDSecondary’s Propagation may be making mistakes (but looks reasonable to me). If you think it can do better for certain cells, then please post an annotated image with how the “declumping” ought to be dividing the cells.