Stopping-time resampling and population genetic inference under coalescent models

Area of Science:

• Computational Biology
• Population Genetics
• Bioinformatics

Background:

• Sophisticated model-based techniques like importance sampling are crucial for DNA sequence data analysis.
• Current methods face limitations in handling large datasets efficiently.
• Stopping-time resampling, introduced by Chen and Liu (2000), enhances importance sampling under specific coalescent models.

Purpose of the Study:

• To develop a generalized framework for designing stopping-time resampling schemes.
• To improve the efficiency and accuracy of analyzing DNA sequence data under various evolutionary models.
• To address the limitations of existing methods in handling large datasets.

Main Methods:

• Development of a novel framework for stopping-time resampling.
• Implementation of the framework on infinite sites and stepwise mutation models.
• Extension of the framework to incorporate crossover recombination.
• Simulation studies to evaluate accuracy and performance.

Main Results:

• The new framework significantly improves the accuracy of likelihood estimation across diverse parameters.
• Direct application of the Chen and Liu (2000) scheme can lead to diminished estimates.
• The proposed method imposes no additional computational burden.
• The framework demonstrates robustness to parameter choices.

Conclusions:

• The developed stopping-time resampling framework offers substantial improvements in DNA sequence data analysis.
• This generalized approach enhances likelihood estimation accuracy compared to previous methods.
• The method is computationally efficient and robust, making it suitable for complex population genetic models.