i am quantifying the amount of (arginine deiminase) ADI gene expressed when bacteria is subject to abundance of arginine (arg+) absence of arginine (arg-), in the availability of light ( light+) absence of light (light-). any suggestions on number of trials? would 5 trials be enough?
as far as i understand u mean by trial the sample like sample size so if thats true there are formulas which can be used to decide the sample size. so there is still need for more information on what and how u measure the ADI gene expression. it would be helpful if u could decribe ur methodand outcome clearly.
I will think about one-way ANOVA if you consider the 4 groups are independent, or two-way ANOVA if you consider there is an interaction between arginine and light. The issue will be how do you interpret the result when you have a significant value of statistics.
One-Way ANOVA is not a good approach, since you have, at least, two interesting factor (abundance x light). There is no doubt about it!
However, there is still one doubt about what "trial" means... If each trial is a group of measures, then you have a Three-Way ANOVA (abundance x light x trial) with repeated mesures at the third factor (trial). If this is not the case and "trial" is just one measure, Daniela Ferreira is right and you should stay on Two-Way. Do not forget to look at interactions!!!
Is there a typo in the table for group C? Did you mean to say "arginine- light-"?
Another question: does each dish or well measured only once? I.e. does each cell in your table represent a measurement taken from a different sample of bacteria?
If the answer to both of the above is "yes", I agree that two-way ANOVA with interactions is the way to go (response is ADI, the predictors are arginine, light, and arginineXlight, and the trials are ordinary replicates). Then, the arginine effect is how much ADI expression changes in bacteria with arginine but without light, the light effect is how much it changes with light but without arginine, and the interaction effect is how much more (or less) ADI expression changes due to having both arginine and light compared to how much it changes due to having just one or the other.
If the answer to the second question is "no" (i.e. some samples were measured more than once), then you have some type of repeated measures design. I would use a mixed-effect model.