Improved methods for mutation rate estimation

In trying to understand the mechanism responsible for somatic hypermutation, one often encounters the problem of accurately estimating the mutation rate. The currently employed method is based on the fluctuation analysis experiment of Luria and Delbrück (1943). While attempting to adapt this method to mutation rate estimation in germinal centers, I found that, due to implicit assumptions about the cell-cycle time distribution, this method underestimates the mutation rate by as much as 30%. In chapter , I introduced a number of methods for estimating mutation rates and constructing confidence intervals, each of which takes into consideration the cell-cycle time distribution. I gave cell-cycle corrections for the mean proportion of mutants in the culture for two of the most common models of cell cycle time. The derivation can be used for other cell cycle time distributions as well. I described a continuum approximation for the Luria-Delbrück distribution, that is considerably easier to use than the currently available methods. I also give a parametrization of this distribution that can be used for cell cultures of arbitrary size, provided that the cell cycle time obeys a shifted exponential distribution. Gamma-distributed cell cycle times do not allow a similar parametrization of cLD. However, I found that the 5% and 95% of the proportion of mutants scale linearly with the mutation rate. Moreover, the slope of these curves does not seem to be sensitive to the culture size. These findings allowed me to design a method for constructing confidence intervals for the mutation rate in this types of cell cultures as well. Finally, I discuss extensions of the above methods for cultures that reach steady-state, as well as for germinal centers. Given that they readily lend themselves to automatization, I believe that the methods that I introduced in this chapter have the potential of becoming widely adopted in the field of experimental biology.