Estimating population genetic parameters and comparing model goodness-of-fit using DNA sequences with error

Area of Science:

• Population Genetics
• Bioinformatics
• Genomics

Background:

• Sequencing errors can introduce bias in estimating evolutionary and population genetic parameters.
• This issue is amplified in deep resequencing studies due to large sample sizes and increased error probability per site.

Purpose of the Study:

• To develop a novel method for simultaneously inferring population mutation rate (theta), exponential growth rate (R), and sequencing error rate (epsilon).
• To evaluate the accuracy of parameter estimation under varying conditions of theta, sample size (n), epsilon, and R.

Main Methods:

• A composite likelihood approach was used to infer population genetic parameters and error rates.
• Simulations were conducted to assess the impact of different parameters on estimation accuracy.
• The maximum composite likelihood estimator (MCLE) for theta was compared with existing methods.
• Parametric bootstrap was employed for model goodness-of-fit testing.

Main Results:

• The proposed MCLE method demonstrates robust performance, particularly with large sample sizes or high error rates.
• The study identified combined effects of theta, n, epsilon, and R on parameter estimation accuracy.
• The MCLE method was successfully applied to real-world sequence data from the ANGPTL4 gene.

Conclusions:

• The composite likelihood method provides a reliable approach for estimating key population genetic parameters and sequencing error rates concurrently.
• This method enhances the accuracy of genetic analyses, especially in deep resequencing projects.
• The findings have implications for understanding genetic variation and evolutionary processes in human populations.