PPNet

Introduction

• What is PPNet? PPNet is designed to uses genome information and analysis of phylogenetic profiles with binary similarity and distance measures to derive large-scale bacterial association networks of a single species.

Installation

PPNet has the following dependencies:

• prokka
• roary
• Python(>=version 3.7)
• Python modules:
  • biopython
  • pyvis
  • numpy
  • scipy
  • statsmodels
  • kneed
  • pyani

• Install with the source codes
  • Download the source codes:

    git clone https://github.com/liyangjie/PPNet.git
    

  • Rename the main program and add the path to the environment variable:

    # Rename PPNet.py to PPNet
    mv PPNet/bin/ppnet.py PPNet/bin/ppnet
    # Give the scripts executable permission
    chmod +x PPNet/bin/*
    # Add the path to the environment variable
    echo export PATH="/Path/to/PPNet/bin:$PATH" >> ~/.bashrc
    source ~/.bashrc
    

  • Install the Python dependencies:

    pip install biopython pyvis numpy scipy statsmodels pyani
    

  • Install the external dependances either from source or from your packaging system:

    prokka roary
    

Usage

ppnet [Options]

Options:
      [-h] show this help message and exit
      [-i1] [Required] The path of input genomes
      [-i2] [Required] The path of phenotype (e.g., pathogenic or non-pathogenic) of all strains
      [-o] The path of output (Default "./PPNet_output")
      [-x] The suffix of genomes data (Default "fasta")
      [-c] number of CPUs to use
      [-a] [Required] Select the algorithm for calculating the correlation coefficient [1-81], or set 0 to use all algorithm.
      [-pt] What percentage of interactions will be visualized (Default "1")

Algorithm

See Algorithm.docx

Examples

ppnet -i1 PATH/to/your/genomes/ -i2 group.csv -x fasta -c 4 -a 1

Input

The genome file should be in fasta format and placed in the same path. The group.csv

Output

• PPNet_output/HQ_data/*: High quality genomes which with N50 > 10000；
• PPNet_output/NR_data/*: Non-redundant genome sets after deduplication；
• PPNet_output/Prokka_result/*: The result files of Prokka
• PPNet_output/Gff_file/*: Include the GFF file extracted from the prokka_result folder with the input file for roary
• PPNet_output/Roary_result/*: Result files generated by roary
• PPNet_output/Roary_result/Statistical_test_result.csv: The result of Fisher's exact test for the distribution of each gene, by default, PPNet reports all genes with a adjusted p-value <0.05.
• PPNet_output/Roary_result/filted_phylogenetic_profile.csv: The phylogenetic profile of orthologs with significantly different distributions.
• PPNet_output/Roary_result/netwrok_result_method_x.csv: List the association coefficient calculated by algorithm x between each pair of genes.
• PPNet_output/Gene_net_x.html: A network plot inferred by algorithm x that can be opened with a browser(Google Chrome,Microsoft Edge etc.).By default, only first percent of interactions were visualized.

License

PPNet is free software under a GPLv3 license.