Co-occurrence-based Horizontal Gene Transfer (CoHGT)

CoHGT (Co-occurrence-based Horizontal Gene Transfer) is a lightweight Python tool designed to identify potential co-transfer events of pathogenicity gene clusters mediated by mobile genetic elements (MGEs) in draft bacterial genomes. It leverages co-occurrence patterns and clustering analysis to infer horizontal gene transfer (HGT) relationships. Features: - Computes Jaccard similarity between annotated gene features across genomes - Supports input of various annotations: virulence genes, resistance genes, insertion sequences (IS), transposases, integrases, plasmid replicons, etc. - Compatible with draft assemblies (no need for closed genomes) Use Cases: - Identifying MGE-mediated gene transfer in pathogen evolution studies - Exploring functional modules in microbial genomics - Investigating the spread of virulence and resistance factors in epidemiological contexts Requirements: - Python >= 3.8 - pandas - numpy Input Format: CoHGT accepts a binary annotation matrix in CSV format, in which each row represents a genome (or isolate) and each column represents a genomic feature (e.g., virulence gene, antimicrobial resistance gene, insertion sequence, integron, plasmid replicon, or other annotated feature). Presence is encoded as 1 and absence as 0. Example: Sample ID, fimH, iroN, IS26, intI1, IncFIB, ... EC001, 1, 1, 1, 0, 1, ... EC002, 0, 1, 0, 1, 0, ... The demonstration data and code is located in the demo. Quick Start: For a single annotation matrix: co_in_1(file, outpath) co_matrix(file, outpath) For two independent annotation matrices: co_two_in_1(file1, file2, outpath) co_for_two(file1, file2, outpath) Example datasets and demonstration scripts are provided in the demo directory. Output The software calculates four commonly used co-occurrence statistics for each feature pair: co-occurrence count, co-occurrence rate, Jaccard index, and phi coefficient Results can be exported as either: long-format tables containing pairwise statistics; or matrix-format files for downstream visualization and clustering analyses. Documentation Further documentation, source code, demonstration data, and future updates are available at: https://ngdc.cncb.ac.cn/biocode/tool/BT007989