Sometimes we see small changes (eg < 30% change) that are significant, but given the inherent variability in the technique I wonder whether I can trust such small changes in Western blot data and whether they are physiologically meaningful?
I think small change like 30% could be meaningful for basal expression.
For phosphorylation/activation of a protein, a twofold increase or decrease is necessary to make sense for me.
Hope that help you
Kevin
Dear David,
My looking on this matter was always as following: biological significance is not fully equivalent to mathematical significance. As a matter of fact, 5% change may be very significant biologically, and sometimes 50% change may be irrelevant (though of course statistically significant!)
I think your question touches the more general issue of data reliability obtained with semi-quantitative methodologies. I would say if your blots are done properly (reproducibly, with large number of replicates, appropriate choice of reference genes, proper normalization of input protein levels, good quality of PAGE and membrane transfer, and using fluorescent instead of chemiluminescent detection), then I would trust even much smaller changes then the 30% in your question. Especially if you can show that they really mean something.
I think Lech is right. If changes are reproducible and depending of course of the loading control protein, the small changes could be important. It depends on the protein of interest that you study and it is important the literature data obtained so far.
Dear David,
I agree with you that this is an important concern and I am often faced to the same question.
For the first part of your question, regarding the fact to "trust or not" the result when the difference is pretty weak, 2 things allow me to trust in the result: First, I think that it is important that the experiment is reproducible, given that western blot experiments are sometimes whimsical. Therefore I am more confident when I reproduce the results twice with 2 independent experiments. Second, I am more confident in the result when it is integrated in a set of results that confirm the interpretation. For example, when I see an increased in the amount of one protein, I am more confident for example if another protein regulated by the same pathway is also increased, or if a protein dowstream the studied protein is also activated, or the ARNm is also increased etc. I like to looking for other complementary experiments that would confirm the Western blot done.
For the second part of your question, regarding the physiological relevance, i think that it is highly dependent of the protein. A small difference is sometimes associated with high physiological consequences, whereas a huge increase can be without any consequence. I think we have to think about the limiting factor of the studied pathway, in order to answer to this point. If your protein is a "rate-limiting" step, you could expect a physiological impact even with a small change in the protein level.
Hope it helps you!
Chears