A new practical guide to the Luria–Delbrück protocol

• The concept of the likely average is obsolete.
• Imperfect plating efficiency can be accounted for by likelihood-based methods.
• Relative fitness measured in a fitness assay can be seamlessly integrated.
• Likelihood ratio tests are feasible even when final cell population sizes differ.
• The R tool rSalvador offers all capabilities discussed in the paper.

Since 2000 several review papers have been published about the analysis of experimental data obtained using the Luria–Delbrück protocol. These timely papers cleared much of the confusion surrounding various methods for estimating or comparing mutation rates. As a result, today the fluctuation test is more widely applied with much improved accuracy. The present paper provides guidelines on a few remaining problems that continue to baffle mutation researchers. Among the issues addressed are incomplete plating, relative fitness, and comparison of experiments where average final cell population sizes differ. It also offers a fresh view on the estimation methods that are based on the sample median.