Using maximum likelihood to estimate population size from temporal changes in allele frequencies

Area of Science:

• Population genetics
• Evolutionary biology
• Quantitative genetics

Background:

• Estimating effective population size is crucial for understanding population dynamics and evolutionary potential.
• Traditional methods, like F-statistic estimators, have limitations in accuracy and bias.
• Temporal changes in allele frequencies offer valuable information about population genetic processes.

Purpose of the Study:

• To develop a maximum-likelihood framework for estimating the number of breeding individuals using temporal allele frequency changes.
• To compare the performance of this new estimator against the F-statistic estimator.
• To extend the framework to detect population growth.

Main Methods:

• Developed a maximum-likelihood framework based on temporal allele frequency shifts.
• Utilized computer simulations to compare estimator performance.
• Extended the model to incorporate exponential population growth.

Main Results:

• The maximum-likelihood estimator demonstrated lower variance and reduced bias compared to the F-statistic estimator.
• The simulations confirmed the improved accuracy of the new method.
• The extended model successfully used temporal allele frequency data from three or more samples to test for population growth.

Conclusions:

• The maximum-likelihood framework provides a more accurate and less biased method for estimating effective population size from temporal allele frequency data.
• This framework offers a robust tool for population genetic studies.
• The ability to test for population growth adds significant value for ecological and evolutionary research.