I know tat ImageJ has options to count cells either manually or automatically but my protein of interest lights up the synapses and hence can't use both. I thought of measuring the mean grayscale value in the area and subtract it from the backgrounds. But then, how accurate is it when I need to compare photos that were taken under various brightness, gamma, white balance and what not! Is subtracting background's grayscale enough a correction?
Hello Vasishta,
You need all the pictures to be captured with the same conditions. Thus avoiding automatic exposure and correction is a must. You will need to setup your best exposure time and keep it the same for all the pictures that will be compared together. One trick is to use the over-exposure tool (that most capture software have) and use an exposure time that do not over-expose the samples that show more signal.
Beyond capturing variability you also have to take in account that developing times when using enzymatic protocols (i.e. Peroxidase-DAB), small variations in tissue thickness among other problems will increase artificial variability.
For me quantification by densitometry (grey-scale mean) of IHC staining is something to avoid. My feeling is that immunofluorescence has a better relation between signal and protein concentration and less variability than enzymatic protocols.
I'd say what David say, plus having a baseline reference (this is, a protein that doesn't change fluorescence intensity) as a control, since you need to be sure that the changes in fluorescence of your protein of interest are because of the variation of that protein, and not due to something global.
Also, you could try to do some biochemistry, purify synapses and perform some western blot analysis.
Hope this helps! Good luck!
- Badges
- Science method