Could someone kindly help me understand the implication of the intensity profile plot between two different tags? E.g EGFP for protein X and mCherry for protein Y.
Thanks.
Remi.
If it is a line plot, you are usually looking at quasi-colocalization in terms of peaks overlapping at the same locations along time line. This does not mean that the proteins are interacting in any way, but can give you an idea if they are associated. Two general cytoplasmic proteins could look very closely associated while having no actual contact in the cell because both will have a peak when a line crosses a cell, and low background outside of a cell.
Very, very close association (R^2 > 0.9) can sometimes mean bleedthrough between channels.
What are you looking for?
Thanks very much for your reply. Yes, it’s a line plot. With reference to this quasi-colocalization, can I, for example, count the number of peaks in line A and line B, and then infer this quasi-colocalization from the number of overlapping peaks?
So, in what situation might a line plot analysis be preferable to actually performing colocalization analysis?
Thanks.
While it is pretty much impossible to prove interaction with a plot like that, it is surprisingly easy to show that two proteins are almost never in the same location at the same time. Or if you want to show that one localizes to the same or different compartments as a third marker (DAPI, MitoTracker, etc). Then you aren’t looking at interactions (which you might want something like FRET or PLA for) so much as broader peaks that coincide with some other feature of the cell. You could try counting overlapping peaks, but you would need to be very careful about providing adequate controls. Two proteins that form peaks but have nothing to do with each other, and each are in 75% of a cell’s cytoplasm might seem very highly correlated by chance. I’m not great with the statistical analysis side of things to help with the specifics, other than being aware that it needs to be accounted for =/
A line drawn through the cytoplasm and nucleus that shows differing peaks for the DAPI containing part of the line versus the non-DAPI areas could support an image showing the overlap. In an obvious example like that, a simple picture would probably also be sufficient, but the line pot might help if you think something is peri-nuclear, and shows a spike right before the DAPI signal gets strongest.
Maybe others will have some better ideas. Oh, and if the interaction is… complicated, you might be able to use the data from the line plot to get an R^2 value. Though that comes with caveats too, since R^2 values are only meaningful for linear relationships.
In the end, if the line isn’t passing through different compartments, or along some kind of axis (through an organoid, for example), you are probably better off defining an area and performing some kind of colocalization analysis, be it Manders or Pearson’s.
Also, as this is less of a microscope question, and more of an analysis question, you might get better or more informative responses on the analysis forum. Similar posts have also been diverted.
3 Likes
Yes, I also recommend you bring this question to the image analysis forum.