Empirical models for substitution in ribosomal RNA

Area of Science:

• Computational Biology
• Bioinformatics
• Molecular Evolution

Background:

• Empirical substitution models are crucial for protein sequence analysis due to large amino acid alphabets.
• Analyzing structured RNA substitutions requires modeling paired bases, increasing parameter complexity.
• Existing models often lack the capacity to fully capture RNA's structural nuances in evolutionary analysis.

Purpose of the Study:

• To develop an empirical method for generating RNA substitution rate matrices that incorporate structural information.
• To create a universal rate matrix for RNA sequences applicable to phylogenetic analyses.
• To enhance the accuracy of evolutionary tree reconstruction for ribosomal RNA.

Main Methods:

• Utilized ribosomal RNA alignments from the European Ribosomal RNA Database.
• Developed a 20-symbol code representing individual bases and 16 ordered base pairs, incorporating secondary structure.
• Estimated evolutionary distances and derived instantaneous and universal rate matrices, integrated into Phylip software for phylogenetic analysis.

Main Results:

• Successfully generated empirical substitution rate matrices for RNA sequences using structural information.
• Empirical models demonstrated good performance on simulated data.
• Produced reliable evolutionary trees for bacterial 16S ribosomal RNA sequences.

Conclusions:

• The developed empirical models effectively capture RNA sequence and structure co-evolution.
• These models are readily implementable and compatible with existing protein sequence analysis tools.
• The method provides a valuable tool for simulating RNA evolution and improving phylogenetic inference.