Data and code for "Replication-driven mutation spectrum asymmetry causes GC-skew in bacterial genomes"

<h2 dir="ltr">Replication-driven mutation spectrum asymmetry causes GC-skew in bacterial genomes</h2><p dir="ltr">This repository contains the raw data, analysis scripts, output results and supplementary materials supporting the manuscript "Replication-driven mutation spectrum asymmetry causes GC-skew in bacterial genomes". The codebase analyzes previously published mutation accumulation experiment (MAE) datasets using a replication strand-specific 12-component mutational spectrum framework.

It evaluates the evolutionary impact of these spectra on bacterial nucleotide composition, including equilibrium base frequencies. Additionally, the nucleotide composition of reference sequences is investigated in relation to replication strand orientation and gene expression levels. Sources for the original MAE datasets and expression data are referenced in the manuscript.</p><h2 dir="ltr">Repository Structure</h2><p dir="ltr">Our workspace is organized into the following directory structure:</p><pre><pre>.<br>├── Scripts/<br>├── Data/<br>│ ├── MAE_data/<br>│ ├── Genbank_data/<br>│ ├── Processed/<br>│ └── Expression_data/<br>└── Results/ <br> ├── Figures/ <br> └── Tables/<br></pre></pre><h3 dir="ltr">Directory Descriptions</h3><ul><li><code><strong>Scripts/</strong></code> — Contains source code and executable scripts used for data processing and analysis.</li><li><code><strong>Data/</strong></code> — Input datasets and intermediate processed files:</li><li><ul><li><code>MAE_data/</code> — Mutational spectrum / MAE (Mutation Accumulation Experiment) raw datasets.</li><li><code>Genbank_data/</code> — Reference genomes and annotation files downloaded from GenBank.</li><li><code>Processed/</code> — Cleaned and preprocessed data ready for downstream analysis.</li><li><code>Expression_data/</code> — E.coli gene expression dataset (e.g., RNA-Seq / TPM values).</li></ul></li><li><code><strong>Results/</strong></code> — Generated output files, tables and figures.</li></ul><h2 dir="ltr">Used software:</h2><ul><li>Python 3.10.12+</li></ul><h2 dir="ltr">Environment and Prerequisites</h2><ul><li><b>Python:</b> <code>>= 3.10.12</code></li><li><b>Environment Manager:</b> <code>venv</code> (standard library) or <code>Conda</code> / <code>Miniconda</code></li><li><b>Build System:</b> <code>pyproject.toml</code> (all required Python libraries are listed in dependencies and installed automatically)</li></ul><h2 dir="ltr">Installation & Setup</h2><ol><li>Clone or extract the project repository and navigate into the root directory.</li><li>Create and activate your preferred virtual environment (e.g., via <code>venv</code> or <code>conda</code>).</li><li>Install the package and all dependencies in editable mode:</li></ol><pre><pre>pip install -e .<br></pre></pre><h2 dir="ltr">Script Execution Workflow</h2><p dir="ltr">To reproduce the analysis, run the scripts in the following order:</p><h3 dir="ltr">Mutation Spectra Analysis</h3><ol><li><code><strong>Raw_MAE_data_standardization.ipynb</strong></code></li><li><ul><li><b>Description:</b> Preprocesses raw MAE supplementary datasets for downstream analyses and extracts strand-specific 12-component mutations using <i>oriC</i> and <i>ter</i> sites.</li></ul></li><li><code><strong>Mutation_spectra.ipynb</strong></code></li><li><ul><li><b>Description:</b> Calculates 6-component and strand-specific 12-component mutation spectra for bacterial species.</li></ul></li><li><code><strong>Permutation_script.ipynb</strong></code></li><li><ul><li><b>Description:</b> Calculates the statistical significance of differences in mutation asymmetry between the leading and lagging strands.</li></ul></li><li><code><strong>Split_mutations_by_strand.ipynb</strong></code></li><li><ul><li><b>Description:</b> Partitions full-genome standardized MAE mutational data into two datasets: genes located on the leading strand and genes located on the lagging strand.</li></ul></li><li><code><strong>Gene_strand_mutspec.ipynb</strong></code></li><li><ul><li><b>Description:</b> Calculates strand-specific 12-component mutation spectra for leading- and lagging-strand genes across bacterial species.</li></ul></li><li><code><strong>Gene_strand_permutation.ipynb</strong></code></li><li><ul><li><b>Description:</b> Calculates the statistical significance of differences in strand-specific mutation spectra between genes located on the leading vs. lagging strands.</li></ul></li><li><code><strong>Mutation_spectra_E_coli_ko.ipynb</strong></code></li><li><ul><li><b>Description:</b> Calculates strand-specific 12-component mutation spectra of repair-defective <i>E. coli</i> knockout strains.</li></ul></li><li><code><strong>Mutspec_count_script.ipynb</strong></code></li><li><ul><li><b>Description:</b> Contains core counting functions from <code>Mutation_spectra.ipynb</code> refactored as reusable modules for <code>Gene_strand_mutspec.ipynb</code> and <code>Mutation_spectra_E_coli_ko.ipynb</code>.</li></ul></li></ol><h3 dir="ltr">Nucleotide & Codon Frequencies Analysis</h3><ol><li><code><strong>Nucleotide_frequencies.ipynb</strong></code></li><li><ul><li><b>Description:</b> Computes nucleotide counts across different genomic positions (four-fold degenerate sites, 1st, 2nd, and 3rd codon positions, and intergenic regions), both strand-specifically and combined relative to the leading replication strand.</li></ul></li><li><code><strong>Equilibrium_nuc_frequencies.ipynb</strong></code></li><li><ul><li><b>Description:</b> Computes expected equilibrium nucleotide frequencies across genomic regions (four-fold degenerate sites, 1st–3rd codon positions, and intergenic regions) derived from mutation spectra.</li></ul></li><li><code><strong>Skew_graphics.ipynb</strong></code></li><li><ul><li><b>Description:</b> Calculates various strand asymmetry metrics (GC-skew, AT-skew, and CT/GA-skew) using output data generated by <code>Nucleotide_frequencies.ipynb</code> and <code>Equilibrium_nuc_frequencies.ipynb</code>, and visualizes the results.</li></ul></li><li><code><strong>Codon_frequencies.ipynb</strong></code></li><li><ul><li><b>Description:</b> Computes normalized strand-specific codon frequencies across four-fold degenerate positions and calculates the inter-strand frequency ratio for each codon type.</li></ul></li></ol><h3 dir="ltr">Expression Data Analysis</h3><ol><li><code><strong>Expression_codon_usage.ipynb</strong></code></li><li><ul><li><b>Description:</b> Computes strand-specific four-fold degenerate codon frequencies for highly and lowly expressed genes, calculating their ratio (high-expression vs. low-expression codon frequency).</li></ul></li><li><code><strong>Expression_clusters_GC_skew.ipynb</strong></code></li><li><ul><li><b>Description:</b> Performs clustering of expression data and calculates GC-skew for each cluster stratified by replication strand and expression group.</li></ul></li><li><code><strong>Expression_permutation_test.ipynb</strong></code></li><li><ul><li><b>Description:</b> Evaluates the statistical significance of strand asymmetry in mutation spectra across high and low gene expression groups.</li></ul></li></ol><h2 dir="ltr">Manuscript Status</h2><p dir="ltr">This repository accompanies the manuscript:</p><blockquote><p dir="ltr"><b>"Replication-driven mutation spectrum asymmetry causes GC-skew in bacterial genomes"</b><br><i>In preparation.</i></p></blockquote><p dir="ltr"><br></p>