A new method for estimating nonsynonymous substitutions and its applications to detecting positive selection

Area of Science:

• Evolutionary biology
• Molecular evolution
• Genomics

Background:

• Standard methods for nonsynonymous substitution (Ka) analysis group all amino acid changes together.
• This approach overlooks significant variations in substitution rates (up to 10-fold).
• Previous attempts to classify changes by physicochemical properties have been ineffective.

Purpose of the Study:

• To propose a novel method for classifying amino acid changes based on evolutionary exchangeability using the Universal index U.
• To introduce the Kh statistic for high-exchangeability amino acid changes.
• To improve the detection of positive selection in evolutionary analyses.

Main Methods:

• Classifying 75 elementary amino acid changes by evolutionary exchangeability using the Universal index U.
• Calculating Ki for each class and deriving Kh for the top 10 classes.
• Applying the Kh estimation method to human-macaque and mouse-rat comparisons.

Main Results:

• The Kh statistic typically accounts for 25-30% of total amino acid changes.
• A 'twofold approximation' was observed: Kh is approximately twice Ka for large substitution numbers.
• The Kh statistic facilitated easier discernment of positive selection signatures compared to Ka.
• Genes with Ka/Ks > 0.5 were identified as having Kh/Ks > 1, indicating adaptive evolution in high-exchangeability groups.

Conclusions:

• The proposed Kh statistic offers a more refined analysis of amino acid substitutions.
• This method enhances the identification of adaptive evolution, particularly in specific amino acid change categories.
• The twofold approximation provides a reliable benchmark for evolutionary rate estimations.