How can I get data for class predictions or bigger studies?
We recommend to use a Reference RNA which should be chosen to have genes expressed in a reasonable way which you would like to study (to get quantitative data and not have a saturated signal or one which is too low), e.g. Stratagene Universal reference RNA (human # 740000; mouse # 740100; rat # 740200). A Reference RNA is a tool to control better experimental variance after RNA preparation. • Compute ratios of your experimental samples with the reference RNA (do not use the housekeeping genes for normalization in the 2nd step as they are likely to be very different between experimental sample and reference sample). • Then decide if the status (not in linear range, qualitative, quantitative / not the significance values of the ratios (unchanged, significant, highly significant) has to be incorporated somehow in the data analysis (like omitting the data which is not quantitative). • Clustering techniques may then be applied and statistical tests on groups like ANOVA.
