RadialCV explanation


#1

Hello, I’m interested in better understanding what RadialCV is measuring (from the output of the MeasureObjectIntensityDistribution module)

The manual states that RadialCV is the “coefficient of variation of intensity within a ring, calculated over 8 slices.”

I understand the coefficient of variation, but what are these 8 slices? How are they calculated, and is there a way to visualize the “slices”?

Is it something like this?

I’m asking because I want to measure how “complete” the expression of a protein is on the plasma membrane. If the membrane is completely covered, the RadialCV should be low. Versus incomplete/partial expression on the membrane which should give a high RadialCV. Does this make sense?

I used the attached custom pipelines, which measure the RadialCV in the membrane (by setting the MeasureObjectIntensityDistribution bin size larger than the ExpandorShrinkObjects size initially used to determine the approximate thickness of the membrane. This was necessary as at least 2 bins were required, but I only needed one. So the second bin measured nothing.)

Determining membrane location


MeasureObjectIntensityDisribution of the membrane

yeastmembrane_small.cppipe (8.7 KB)
yeastmembrane_large.cppipe (8.7 KB)


#2

If you’ll pardon my utterly awful drawing skills, RadialCV lets you know how much variability there is between 1A-1H (or 2A-2H, etc, depending on which ring you’re looking at). FractionAtDistance, which you graphed, lets you know how much is in ring 1 vs ring 2, etc.

RadialCV is therefore a decent metric to choose, though you may want to figure out a way to pre-filter the cells that are budded or aberrantly split or merged, since those will have a high RacialCV as well.


#3

Thanks for the explanation and detailed drawing. This is more or less how I was picturing it, so I really appreciate the clarification!