A Probabilistic Model for Indel Evolution: Differentiating Insertions from Deletions.

Area of Science:

• Molecular Evolution
• Computational Biology
• Bioinformatics

Background:

• Insertions and deletions (indels) are fundamental to molecular evolution but are computationally complex to model.
• Existing probabilistic models often oversimplify indel evolution by assuming equal rates and length distributions for insertions and deletions.

Purpose of the Study:

• To develop a more sophisticated and computationally feasible probabilistic model for indel evolution.
• To improve parameter estimation for indel models using approximate Bayesian computation.
• To address biases in indel parameter inference caused by sequence alignment programs.

Main Methods:

• Proposed a richer indel model distinguishing insertion and deletion rates and length distributions.
• Introduced novel summary statistics for approximate Bayesian computation-based parameter estimation.
• Developed a method to correct for alignment-induced biases in empirical data.
• Implemented a model-selection framework to compare the richer model against simpler ones.

Main Results:

• The proposed inference scheme and model-selection procedure demonstrated high accuracy on simulated data.
• The richer indel model provided a better fit to a large number of empirical data sets.
• Analysis revealed that deletion rates typically exceed insertion rates across most empirical data sets.

Conclusions:

• The developed probabilistic model and inference methods offer significant improvements for studying indel evolution.
• Distinguishing between insertion and deletion processes is crucial for accurate evolutionary modeling.
• Empirical data suggests a higher rate of deletions compared to insertions in molecular evolution.