DLS sizing for separations
I’ve updated by 3D separations figure by using actual DLS sizes rather than the dimensions roughly obtained from crystal structures. I have run into a snag with reporting the data however – use sizes by intensity or sizes by number. The difference is due to the fact that the raw DLS data is from intensity, and Mie theory indicates that larger particles have larger intensities. The software makes a correction to give the number data that shifts distributions toward the smaller sizes.
In the past we have used the number data instead of intensity. A few of the references I’ve seen and even the software itself seems to prefer the intensity data. For the larger proteins, the intensity data seems better, but the smaller proteins may be a bit bigger than expected using this same data. Remember that DLS assumes a sphere and is not always the most accurate method for size determination. Here I present the results in both forms:
Another consideration: I am reporting the size of the DLS peaks for 5 repeats, and the error is the standard deviation of these peaks. A better error measurement would take into account the spread of these peaks.
UPDATE:
Redo of the chart that averages literature values with DLS peaks and uses the range between the two for the error bars:
