Interpretation of Output - What is a "true or real" difference? (Jul/24/2007 )
I have been running Sybr Green based real time experiments. I use the deltaCt method to calculate differences between my treatment samples and control after normalizing with a housekeeping gene. My problem arises in interpretation of results. I see very small differences for my delta Ct i.e. 2.5 or 2.8 or 2.2. Is this really a large enough difference for me to call a difference? I ask because when I look at literature I feel like people have 10 or 20 fold-differences and I look at my 2.5 and think this is not much of anything so should I even repeat it. Any thoughts on what you feel the delta Ct should be before I claim differences? Or can you refer to any references that discuss what is a "real or true" difference is?
well, you must use statistical analysis on your results and calculate the p-value to determine statistical significance of the increase
please also remember - an increase in Ct of 2.5 doesn't mean twice as much, it means 2^2.5 - or ~6 times as much - message in the cell. if you take one value in the cell and increase it by 6 times, don't you think it could be important?
the main question is statistical significance, especially if your only data is from real-time. if your data is also supporting increase in actual protein expression (shown by flow, WB, IP, whatever) then a 6-fold difference can be very important.