q-PCR statistical analysis - (May/26/2013 )
I have quaestion regarding the statistical analysis of the real time PCR results.
I have DBA , BL6 mice and I measured the expression level of 10 miRNAs in each mice in two time points 24 and 498 hrs. Now I compare between the 24 hrs. and 48 hrs expression level of each miRNA with regard to the 0 point level. I have only one value for the fold change at 24 and 48 hrs. So how can I perform statistical analysis with only one value. Always in these experiment I have replicate in the CT but with fold change I have only one value.
My questions :
1. What is the suitable test to be used if I want to say that , for example for one miRNAs, that it has a significant difference at 24 hrs between DBA and BL6 mice. I think unpaired T test , here is nice to be performed but two problem facing me. the first that I have to create a graph with the fold change ( not the CT values) and I have only one fold change value, the thing that is unacceptable for program like prism as it need replicates. The second, that I have 10 miRNAs and I need to create one graph for all of them together due to space limitation.
I think you really need to have biological not just technical repeats to derive statistical significance, otherwise you results won't have any statistical power to prove your hypothesis. To make things simple, let's just look at one miRNA and one mouse strain. You want to know its expression changes at different time points in the same strain. The appropriate biological repeats will be measuring its expression in several mice of the same age and genotype. Is that possible? Also the number of mice required for each time point has to be determined by calculating a sample size based on the inter-mouse variation in measurement values you get.