SCUMBLE: a method for systematic and accurate detection of codon usage bias by maximum likelihood estimation

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• The genetic code is degenerate, with multiple codons encoding the same amino acid.
• Synonymous codon usage varies significantly across species and genes, a phenomenon known as codon usage bias.
• Existing methods explain only a fraction of observed codon usage variation.

Purpose of the Study:

• To develop a novel model for codon usage bias inspired by statistical physics.
• To identify and quantify different sources of codon bias in genomes.
• To improve the understanding of factors influencing gene expression.

Main Methods:

• Developed an explicit model of codon usage bias using principles from statistical physics.
• Integrated the model with a maximum likelihood approach for data analysis.
• Applied the algorithm to Saccharomyces cerevisiae and 325 prokaryotic genomes.

Main Results:

• The statistical physics-based model effectively explains codon usage variation.
• The approach clearly identifies diverse sources contributing to codon bias.
• The model accounts for essentially all observed variance in most tested genomes.

Conclusions:

• The new model provides a powerful framework for understanding codon usage bias.
• This approach offers a more comprehensive explanation for codon bias than previous methods.
• The findings have implications for gene expression regulation and synthetic biology.