What is the best statisical analysis to use? - (Apr/04/2014 )

Hi

I ran an experiment to measure the effect of two drugs combined on cell apoptosis.  I used 8 different concentrations of the first  drug alone, and 8 different concentrations of the first drug combined with a constant concentration of the second.  I measured cell viability using the MMT method.

Which ANOVA analysis would be best to use?  I thought two way as I wanted to measure the effect at each concentration and compare it to each concentration of the first drug alone.

It was suggested to me to use one way as I am only measuring one factor, but I consider it to be two. 

Any help would be appreciated.

What you are looking to compare is two dose response curves so ideally you would look to compare the (logistic) parameters that describe that response; but the estimation of the errors surrounding non-linear parameters can be tricky. A graph with error bars at each concentration may be all the evidence you need.

As a fall back, you have designed the trial as a form of paired test at each of the 8 concentrations of the first drug, so it is possible to test the significance of the difference between the presence or absence of the second drug. Given the non-linearity perhaps the best start point is a non-parametric test.

As a final comment, be careful not to get caught into claims of synergy. At the very least you need a dose response for the second drug on its own.

good luck