Code for plasmid taxonomic unit (PTU) connectivity and genetic nearest-neighbour analyses

Abstract

OverviewThis repository contains the R and Python scripts developed to investigate genetic connectivity and nearest-neighbour relationships among rare and common plasmid taxonomic units (PTUs). The code supports the reproducibility of three complementary analyses performed in the associated research study.1. PTU connectivity analysisAn R script evaluates genetic connectivity between rare and common PTUs using 100 size-matched subsampling iterations. In each iteration, 300 rare and 300 common PTUs are selected, with plasmid counts matched between groups. Connectivity is quantified by summing normalized inter-PTU edge weights across three categories: common–common, common–rare, and rare–rare. Statistical differences are evaluated using a Friedman test and paired Wilcoxon signed-rank tests with Benjamini–Hochberg correction. The script generates Figure 4A in PNG and SVG formats, together with the underlying numerical results and statistical summaries.2. Whole-plasmid genetic nearest-neighbour analysisA Python script identifies the genetically closest different PTUs using whole-plasmid average nucleotide identity (ANI) and symmetric aligned fraction (AF). Qualifying comparisons require ANI ≥ 70% and AF ≥ 50%. Directional similarities are aggregated into symmetric PTU-level scores to identify nearest genetic neighbours. A permutation-based analysis with 10,000 iterations evaluates the representation of common PTUs among the nearest neighbours of rare PTUs.3. Relaxase phylogenetic nearest-neighbour analysisA Python script examines evolutionary relationships among PTUs using previously inferred relaxase phylogenies from the MOBP1, MOBF, MOBV, and MOBQ families. Patristic distances are used to identify the closest different PTUs, accounting for tied neighbours and heterogeneous classifications across sequence clusters. Statistical significance is assessed using 10,000 permutations of rare/common PTU labels within each relaxase family.ReproducibilityThe repository includes the analysis scripts, documentation, and specifications of the required input files. Fixed random seeds are used to support reproducibility. Instructions for running each analysis, software dependencies, and descriptions of generated outputs are provided in the accompanying README file.