Sample size determination for the fluctuation experiment

- Sample size determination for the fluctuation experiment so far relies on intuition.
- The expected Fisher information can be utilized to help determine sample size.
- Two criteria are suggested to guide the determination of sample size.
- An explicit formula is given to calculate the required sample size.

The Luria-Delbrück fluctuation experiment protocol is increasingly employed to determine microbial mutation rates in the laboratory. An important question raised at the planning stage is “How many cultures are needed?” For over 70 years sample sizes have been determined either by intuition or by following published examples where sample sizes were chosen intuitively. This paper proposes a practical method for determining the sample size. The proposed method relies on existing algorithms for computing the expected Fisher information under two commonly used mutant distributions. The role of partial plating in reducing sample size is discussed.