An almost infinite sites model.

Area of Science:

• Molecular Evolution
• Computational Biology
• Population Genetics

Background:

• Accurate molecular evolution models are crucial for understanding genetic variation.
• The infinite sites model is computationally tractable but ignores recurrent mutations.
• The finite sites model accounts for recurrent mutations but is computationally infeasible.

Purpose of the Study:

• To develop a novel molecular evolution model that is both computationally tractable and allows for recurrent mutations.
• To bridge the gap between the infinite sites and finite sites models.
• To provide a flexible mutation model for analyzing genetic variation.

Main Methods:

• Developed the almost infinite sites model (AISM).
• Provided a recursive characterization of the model's likelihood under complete linkage.
• Outlined a parsimonious approximation scheme for likelihood computation.

Main Results:

• The AISM admits recurrent mutations and is computationally tractable.
• Demonstrated the model's utility with simulated and human mitochondrial data.
• Showed that AISM accurately approximates the maximum likelihood estimator of the mutation rate when combined with constraints on mutation events.

Conclusions:

• The almost infinite sites model offers a computationally efficient and flexible approach to molecular evolution.
• AISM successfully incorporates recurrent mutations, addressing limitations of the infinite sites model.
• The developed model has practical applications in analyzing genetic variation and estimating mutation rates.