the importance of quantitative analysis

"In practice, FRET can be detected when the distance between the dyes is < 2 x Ro (27).  The Forster radius for the Alexa dyes used was calculated to be 5.4 nm (see Materials and Methods).  This means that the distance between PLA2 molecules at the end of the lag phase and the burst of activity are, on average, below 11 nm."

 - "Amyloid-Type Fiber Formation in Control of Enzyme Action: Interfacial Activation of Phospholipase A2", Code et al., Biophysical Journal 95:215 (July 2008)


The really great thing about FRET (Forster resonance energy transfer) is its high signal-to-noise ratio: you get a signal when two molecules are close, and you don't when they're far.

So I have no idea how they can conclude, from the fact that they have a FRET signal, that the distance "on average" is below 11 nm: it seems to me they can only conclude that at least some of the molecules are within 11 nm of each other.  Any determination of what proportion of the molecules are within 11 nm would require a more quantitative analysis comparing the amplitude of the signal they detect with the amplitude they would expect if all their molecules were lit up.  And, because you aren't seeing the molecules that are far away from one another, I have no idea how you would determine the average distance.

They go on:

"On the other hand, for a uniform distribution the average distance between PLA2 molecules on the liposome surface would be ~30 nm (Data S1).  Thus, the sole fact that we see FRET demonstrates a non-uniform distribution of PLA2 on the membrane surface."

I guess this statement is true: you would indeed expect zero FRET signal if all the molecules were exactly 30 nm away from one another.  But the statement still seems vacuous to me, because when do you ever expect a uniform distribution in that sense?

Consider: if I drop 100 balls onto a 10 x 10 grid, and if the balls don't interact with one another in any way (imagine the squares are very large compared to the size of the ball, for example), although you do indeed expect to find, on average, one ball per square, I would actually be quite surprised if I found exactly one ball per square.  Instead, I would expect some squares to have more than one ball in them, and some squares to have no balls in them.  Considering from the balls' point of view, I would expect some balls would be close to other balls, and some balls would be far from other balls.

Similarly, if molecules are randomly distributed on the surface of a liposome, I would expect some molecules, by chance, to be close to one another, and some, by chance, to be far away from any other molecules.  If I plotted the distance between molecules as a histogram, I would expect a normal distribution around 30 nm (assuming neither attraction nor repulsion among molecules, our null hypothesis).

So, once again, because FRET will give a signal for close molecules, but no signal for molecules far away from one another, even with this random distribution, I would still expect some nonzero FRET signal.  So once again, in order to determine whether there are additional forces at play here, I would still think that "the sole fact that we see FRET" is not the evidence we need: what is needed here is a quantitative analysis comparing the signal observed to the signal expected.


I once again conclude that the reviewers were not doing their job.  Or am I missing something here?