I have received the data from RT-qPCR and I want to start a statistical analysis. The first step is the normality check. I have 2 diferrent plant tissues (close related,same family), one gene of interest and one reference gene (in one RT-qPCR). A normality check will be done in all samples and in Cq values or in every plant tissue and gene (4 comparisons). It is the first time I do something like this,sorry for the question.
Actually, checking the assumptions is rather the last step of your analysis.
The dCts values should be considered the actual measurements of the qrt-PCR. These are (very typically) the values you will analyze regarding mean differences between your tissues (or experimental conditions).
The typical way is to first fit some linear model representing your study design and estimate the relevant coefficient; in your case this would be the difference in [mean] dCt values between the tissues. The fitted model will then predict the expected dCt for each tissue, and the differences between the observed values and the predicted values are called the residuals. These residuals should resemble a reasonable sample from a normal distribution, and this is what you can check eventually using residual diagnostic plots. The normal-quantile-quantile plot may be the most relevant plot in your case, but a plot of the residuals versus tissue (or versus the predicted value) is probably a good additional check.
- Badges
- Science topic