Error bars in Multi-plate analysis - (Jan/17/2008 )

Hello all
I’m trying to finish the analysis of my Real-Time PCR data and I’m stuck in the last graph.

How do you do when you have done multiple plates and you want to represent the average of your result? The problem is that if I just calculate the Mean and the SD I have the feeling that is not a real representation of the data.

Maybe is easy to understand me with and example:
Gene A Run 1: Ratio A/ Reference is 3.45 SD 0.7
Gene A Run 2: Ratio A/Reference is 3.5 SD 0.8
If I do just MEAN and SD, I have 3.475 SD 0.035
I think some how in the calculation the SD from Run 1 and 2 should be include….. But how?

Thanks
Maribel

---

i think SD should be like this
for example
Gene A Run 1: Ratio A/ Reference is 3.45
Gene A Run 2: Ratio A/ Reference is 3.5
Gene A Run 3: Ratio A/ Reference is 3.55
so the average is 3.5 and SD is 0.05

---

Yes, I think this is mathematically correct but may not reflect the reality. (At least the reality of the experiment
As I explain before in the way you suggested it looks like an experiment with a very tight SD, but that MEAN could come from 3 different runs with big SD in each of the runs.

I think that the ABI 7500 has the function of multiplate analysis. (I was using before that machine but I never bother to look at the way the data come out of the analysis).
If any of you is using that machine and doing multiplate, could please give me an example of the results after the software does the analysis?

Thanks again!

---