Exercise 2: A Large Study
TODO Convert this to h5p
In my qPCR experiment I want to study the pattern of expression of 96 genes (genes of interest and reference genes) in 96 samples of interest, divided into a few groups. I want to use 2 PCR replicates for each reaction.
| Do I need to include IRCs (inter-run calibrators) ? |
|---|
| No, I can fit all samples of the same gene on the same plate so I don’t need to include IRCs. |
I want to include PCR replicates.
| Do I need to include IRCs when I work on a 96 well plate ? |
|---|
| Yes, I have 192 reactions per gene so I cannot place them on the same plate. Remember that replicates have to be located on the same plate ! |
| Do I need to include IRCs when I work on a 384 well plate ? |
|---|
| No, I have 192 reactions per gene so I can even place two genes on the same plate. |
I want to include no template controls but I don’t want to increase the
number of plates.
| What is the most elegant strategy to make room for including negative controls ? |
|---|
| This kind of study screen for expression patterns and requires statistical analysis. Since you have many samples divided over a few groups it means you have many biological replicates so you could easily do without the PCR replicates. By doing so you preserve the biological variability which is often far greater than the technical variation. |