I am doing Western blot data analysis where I have images from a number of experiments (where the samples in the experiments are biological/technical replicates). For each experiment, I labelled the membrane with antibodies against the same proteins of interest.

I want to combine the data from these experiments so that they are all displayed on the same graph.

I took my images at a variety of exposure times (e.g. 5 sec, 30 sec, 1 min, etc). For one protein X, I have found that the images taken at 30 seconds exposure are the best for analysis, as the bands are saturated at higher exposure times.

However, for one experiment, for the protein X labelling, I cannot use the 30 second exposure image for my analysis as there are ghost bands. I have an image taken at 5 seconds exposure which I can use instead.

I was wondering when combining data from western blot images from different experiments, for a specific protein of interest (e.g. protein X), do all the images have to be taken at the same exposure time? All the protein X images I will be using were taken at 30 seconds exposure except for one which is taken at 5 seconds exposure.

Any insights are appreciated.


Generally no, you can't directly compare them and you can't easily compare across gels. I have done many hundreds of western blots, a large proportion of which were for quantitation, and in general quantitation on western blots is difficult (because of inconsistency run to run) and not highly quantitative - a 2-fold change is detectable and probably real, but a 10% change - unlikely to be real.

If you want a real example of how to compare and observe this (works with the Bio-Rad system or others where the transfer is in the same space) - run exactly the same amount of protein on two identical gels, place them both into the same transfer tank. Compare protein stain on the membranes after transfer. The one closest to the black terminal will have stronger staining = better transfer.

However, what you can do is convert the numbers you get into a relative expression on each of the blots and then use those numbers for comparison.

For example you compare expression of protein X to a normalizer (often B-Actin) and get fold-change data out of that, or you can compare Sample X untreated to treatments A, B, C and get data out that way.

  • $\begingroup$ Thank you for your reply. Just to clarify, I can compare the images at different exposures, but only after normalizing the target protein signal to an internal loading control? $\endgroup$ – ceno980 May 18 at 2:46
  • $\begingroup$ Correct. In fact you need to do this for every blot that you run. You cannot pool results from different membranes otherwise, even if all the conditions under which they were run and probed are the same. $\endgroup$ – bob1 May 18 at 3:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.