How many XFe-96 cartridges ?

In the updated XFe-96 cartridge well plate set-up below, we can compare between the age groups and genders between healthy controls and Parkinson's patients for study 1, as well as running a duplicate on the same plate.

This offers the advantage of both initial experiment and confirmation run being done under the same experimental conditions. 

One of the issues yet to be understood is the expectations for the number of runs required due to the variability or accuracy of the XFe-96 results, ie how many technical references are needed for this type of experiment - is one duplicate sufficient or do we need to run another plate to increase the testing from two cases to four cases ?

The good news is that if we need further plates the costs seem to be affordable if we assume we can buy the extra cartridges from the Binger laboratory rather than a bulk order from Agilent

How many patients are significant ?

In order to calculate an effective sample size that will yield a significant result I have set up Gpower using the Linear Multiple Regression: Fixed model single regression co-efficient with the following parameters:

1. Confidence interval of 95%
2. Critical value of 0.05%
3. Power of the test at 80%, and 
4. Large Effect size of 0.80
5. Number of predictor variables set at 1

The analysis results in a proposed sample size of 10 which seems close to the 10 sample we are expecting to get access to from the hospital.