how I can prepare data for heat-map?

Question

[mirzahosseini.z]

I want to show my proteomics data (protein abundance) by heat-map. I have 1 control and 4 treatment conditions. But I don't know how exactly I can prepare data for heatmap. Actualy, I've calculated fold changes for each sample by dividing protein abundance of that sample to protein abundance of correlated control. Then I calculated log 2 for fold changes. Is it the right way to calculate fold change? If yes, the heatmap shouldn't have control, and it will be just for representing treatments. Right? After clustering I don't know how I can interpret it. Is it enough to represent just similarity by clustering? Or I am expected to eliminate or change some of my data according to the result of clustering?

Thanks in advance Zahra

Answer 1

[Devon Ryan]

Normally one plots either the intensity values or some variant of log-transformed intensity values. Don't make a heatmap of fold-changes, that's largely worthless. You don't have to do any clustering at all. If you'd like to, then you might see how your treatment replicates cluster and see if you have any proteins that co-vary.