Codon Usage Bias Analysis

A suite of functions for rapid and flexible analysis of codon usage bias. It provides in-depth analysis at the codon level, including relative synonymous codon usage (RSCU), tRNA weight calculations, machine learning predictions for optimal or preferred codons, and visualization of codon-anticodon pairing. Additionally, it can calculate various gene- specific codon indices such as codon adaptation index (CAI), effective number of codons (ENC), fraction of optimal codons (Fop), tRNA adaptation index (tAI), mean codon stabilization coefficients (CSCg), and GC contents (GC/GC3s/GC4d). It also supports both standard and non-standard genetic code tables found in NCBI, as well as custom genetic code tables.

cubar

Comprehensive Codon Usage Bias Analysis in R

Table of Contents

• Overview
• Features
  • 🧬 Codon-Level Analysis
  • 📊 Gene-Level Metrics
  • 🛠️ Utilities & Tools
• Why Choose cubar?
• Installation
  • Stable Release (Recommended)
  • Development Version
  • Dependencies
• Documentation & Tutorials
  • 🎯 Getting Started
  • 📚 Advanced Topics
• Example Workflow
• 🆘 Getting Help
• Related Packages
• License
• Acknowledgments

Overview

Codon usage bias refers to the non-uniform usage of synonymous codons (codons that encode the same amino acid) across different organisms, genes, and functional categories. cubar is a comprehensive R package for analyzing codon usage bias in coding sequences. It provides a unified framework for calculating established codon usage metrics, conducting sliding-window analyses or differential usage analyses, and optimizing sequences for heterologous expression.

Features

🧬 Codon-Level Analysis

• RSCU calculation: Relative synonymous codon usage analysis
• Amino acid usage: Frequency of each amino acid in sequences
• Codon weights: Calculate weights based on gene expression, tRNA availability, and mRNA stability
• Optimal codon inference: Machine learning-based identification of optimal codons
• Codon-anticodon visualization: Visualization of codon-tRNA pairing relationships

📊 Gene-Level Metrics

• Codon frequency tabulation: Count codon occurrences across sequences
• CAI (Codon Adaptation Index): Measure similarity to highly expressed genes
• ENC (Effective Number of Codons): Assess codon usage bias strength
• Fop (Fraction of Optimal codons): Calculate proportion of optimal codons
• tAI (tRNA Adaptation Index): Match codon usage to tRNA availability
• CSCg (Codon Stabilization Coefficients): Quantify mRNA stability effects
• Dp (Deviation from Proportionality): Analyze virus-host codon usage relationships
• GC content metrics: Overall GC, GC3s (3rd codon positions), GC4d (4-fold degenerate sites)

🛠️ Utilities & Tools

• Sliding window analysis: Positional codon usage patterns within genes
• Sequence optimization: Redesign sequences for optimal expression
• Differential codon usage: Statistical comparison between sequence sets
• Quality control: Comprehensive CDS validation and preprocessing

Why Choose cubar?

• 🚀 High Performance: Process large datasets (>100,000 sequences) efficiently using optimized Biostrings and data.table backends
• 🧬 Flexible Genetic Codes: Support for all NCBI genetic codes plus custom genetic code tables
• 🔗 R Ecosystem Integration: Seamlessly integrate with other bioinformatics and data analysis packages
• 📚 Comprehensive Documentation: Extensive tutorials, examples, and theoretical background
• 🔬 Research Ready: Implements established metrics with proper citations and validation

Installation

Stable Release (Recommended)

Install the latest stable version from CRAN:

install.packages("cubar")

Development Version

Install the latest development version from GitHub:

# Install devtools if not already installed
if (!requireNamespace("devtools", quietly = TRUE)) {
    install.packages("devtools")
}

# Install cubar from GitHub
devtools::install_github("mt1022/cubar", dependencies = TRUE)

Dependencies

System Requirements:

• R (≥ 4.1.0)

Required Packages:

• Biostrings (≥ 2.60.0) - Bioconductor package for sequence manipulation
• IRanges (≥ 2.34.0) - Bioconductor infrastructure for range operations
• data.table (≥ 1.14.0) - High-performance data manipulation
• ggplot2 (≥ 3.3.5) - Data visualization
• rlang (≥ 0.4.11) - Language tools

Note: Bioconductor packages will be installed automatically, but you may need to update your R installation if you encounter compatibility issues.

Documentation & Tutorials

📖 Complete documentation is available within R (?function_name) and on our package website.

🎯 Getting Started

• Introduction to cubar - Basic usage and core functionality
• Non-standard Genetic Codes - Working with alternative genetic codes
• Codon Optimization - Sequence optimization strategies

📚 Advanced Topics

• Mathematical Foundations - Detailed theory behind the metrics
• Function Reference - Complete function documentation

Example Workflow

Here's a typical analysis workflow demonstrating key functionality:

library(cubar)
library(ggplot2)

# 1. Load and quality-check sequences
data(yeast_cds)
clean_cds <- check_cds(yeast_cds)

# 2. Calculate codon frequencies
codon_freq <- count_codons(clean_cds)

# 3. Calculate multiple metrics
enc <- get_enc(codon_freq)           # Effective number of codons
gc3s <- get_gc3s(codon_freq)         # GC content at 3rd positions

# 4. Analyze highly expressed genes
data(yeast_exp)
yeast_exp <- yeast_exp[yeast_exp$gene_id %in% rownames(codon_freq), ]
high_expr <- head(yeast_exp[order(-yeast_exp$fpkm), ], 500)
rscu_high <- est_rscu(codon_freq[high_expr$gene_id, ])
cai <- get_cai(codon_freq, rscu_high)

# 5. Visualize results
df <- data.frame(ENC = enc, CAI = cai, GC3s = gc3s)
ggplot(df, aes(color = GC3s, x = ENC, y = CAI)) + 
  geom_point(alpha = 0.6) + 
  scale_color_viridis_c() +
  labs(title = "Codon Usage Bias Relationships",
       x = "Effective Number of Codons", y = "Codon Adaptation Index")

🆘 Getting Help

• 📋 GitHub Issues: Report bugs, request features, or ask questions
• 📖 Documentation: Check function help (?function_name) and online docs

Related Packages

For complementary analysis, consider these R packages:

• Biostrings - Sequence input/output and manipulation
• Peptides - Peptide and protein property calculations

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• GitHub Copilot was used to suggest code snippets during development
• GitHub Education for providing free access to development tools
• The R and Bioconductor communities for excellent foundational packages
• Contributors and users who have provided feedback and improvements