cross-biome-microbial-networks

Accompanying data and source code for the manuscript https://doi.org/10.1101/2022.09.11.507163

The different subdirectories contain:

• 00-algorithm: source code for our network generation algorithm, and a raw mySQL dump of the database used to perform the study
• 01-pathways-per-genus: source code for the pathway inference algorithm
• 02-network-generation: commands and auxiliary data used for network generation and annotation
  • 02-network-generation/01-per-environment-networks.sh: commands used to generate the individual per-environment networks
  • 02-network-generation/01-output: output of running the commands in 02-network-generation/01-per-environment-networks.sh. For each environment, this includes a presence-absence matrix of genera in samples, and the resulting networks in the gpickle and xml formats
  • 02-network-generation/02-merge-and-annotate.sh: commands used to combine the individual environmental networks into a multi-environment network, and perform phylogenetic and functional annotation
  • 02-network-generation/02-output: output of running the commands in 02-network-generation/02-merge-and-annotate.sh
    • 02-network-generation/02-output/merged6.2.unlooped.avgFunPhyl.pathways.xml: final combined and annotated network
    • 02-network-generation/02-output/merged6.2.networkTable.csv: network table containing annotations for each node in the network
    • 02-network-generation/02-output/consensusNet.sif: consensus network (edge support > 70) in the SIF format
  • 02-network-generation/goodMetaCyc: accompanying data
    • 02-network-generation/goodMetaCyc/oct2020.combined_noPWY0-1324.tsv: fraction of genomes from each genera containing any given pathway
    • 02-network-generation/goodMetaCyc/phylodist_clean.table.tsv: phylogenetic distances between genera
    • Other accessory files linking MetaCyc pathway IDs to pathway names and broader functional categories
• 03-analysis: R code used to analyze the results, and the resulting figures

The following conda environment should provide the libraries required to run the different steps of the analysis: conda create -c conda-forge -c r -n microbialNetworks networkx==1.11 lxml pandas scipy rpy2 mysqlclient cython r r-ade4 r-ggplot2 r-reshape2 r-purrr r-gplots r-dendextend r-svglite r-stringr