A Statistical Guide to the Design of Deep Mutational Scanning Experiments

Area of Science:

• Evolutionary biology
• Molecular biology
• Genetics

Background:

• Characterizing the distribution of mutational effects is crucial in evolutionary biology.
• Deep-sequencing enables simultaneous estimation of fitness effects for numerous engineered mutations via bulk competition experiments.
• Experimental design factors influence the resolution of fitness effect estimations.

Purpose of the Study:

• To provide guidelines for optimizing time-sampled deep-sequencing bulk competition experiments.
• To identify key parameters for improving the precision and power of fitness effect measurements.
• To develop a tool for quantifying experimental precision and identifying deviations from neutrality.

Main Methods:

• Analytical approximations and simulations were used to model bulk competition experiments.
• The study focused on optimizing the number of mutants, sequencing depth, and number of sampled time points.
• A formula for calculating the 95%-confidence interval for measurement error was derived and implemented in a web tool.

Main Results:

• Sampling more time points and extending experiment duration disproportionately improves precision compared to increasing sequencing depth or reducing mutant numbers.
• Clustering time points at the beginning and end of a fixed-duration experiment enhances power for assessing selection coefficients.
• The developed formula and web tool allow for a priori precision quantification and statistical threshold determination for neutrality deviations.

Conclusions:

• Strategic temporal sampling is key to maximizing precision in deep-sequencing competition experiments.
• The findings offer practical guidance for designing more powerful and accurate evolutionary studies.
• The interactive tool aids researchers in optimizing experimental setups and interpreting results for selection coefficient estimation.