Co localization analysis


I have stained for BDNF and p-MAPK (phosphorylated-MAPK). BDNF is a ligand and when it binds to TrkB(receptor), there are multiple down stream process.
p-MAPK is one of the downstream protein of BDNF-TrkB interaction.
Now, I want to do a correlation analysis between BDNF and p-MAPK and I do not want to proceed with co-localization because these 2 proteins don’t really overlap as p-MAPK is a downstream protein. Are you aware of any correlation software that would help us in this analysis. Also, is there any other useful analysis that gives us the information about their interaction.

Thank you

Can you explain what kind of correlation you’re looking for?

E.g. do you want to test if the proteins are likely to be in the same cell, or in neighboring cells?


Perfect guess! I want to test if they are likely to be in the same cell. So, is there a way to know that?

Thanks for your response.

The approach is going to depend on the distribution you see, and on how software can define “the same cell.”

Actually it would be best if you could post an image to illustrate.

To answer your original question, classical correlation analysis describes the relationship between the intensity of the 2 channels. You’re looking for a correlation between the position of the bright areas. There might be a correlation analysis that would work but I’m not experienced in that area.

But you might be able to do it in other ways. One classical way is to define individual areas (cells) positive for one label (such as p-MAPK) and then measure the amount of the other label (BDNF) within a certain radius.

It will help if you can articulate for yourself how you would define this colocalization, or lack thereof. Then you will have an easier time translating the problem into an image analysis workflow.

Hope this helps.

1 Like