metagenome-pipeline

This repository stores the codes to reproduce the analysis in Fujimoto et al. 2024. software_version.list.txt lists the version of software used in the analysis. The analysis was conducted on SHIROKANE OS5.

The pipeline codes are written by Dr. Yasumasa Kimura.

Usage

The codes are intended to be run on SHIROKANE Supercomputer environment. Each script submits the necessary jobs by qsub in each step per sample. inifile package must be installed by gem install inifile. The dir_home_ variable in each script must be set to the pipeline directory holding the scripts directory and a user-defined profile file specifying the environmental variables while running the script. Also, the software directory (e.g. cutadapt_dir) should be set to the path to the software.

Two configuration files must be made to specify the sample and project information as below.

• info.raw_data.cfg

shell [run1] project = (Project title) dir_project_ = (Absolute path to the project directory) dir_run_ = (Absolute path to the sequence directory [e.g. directory produced after HiSeq CASAVA software]) samples = (Comma-separated sample names)

shell [run1] project = Project1 dir_project_ = /home/user/Project1 dir_run_ = /home/user/Project1/fastq samples = SAMPLE001,SAMPLE002,SAMPLE003

• setting.cfg

shell [project1] project = (Project title) type = (virome OR bacteriome) dir_project_ = (Absolute path to the project directory) samples = (Comma-separated sample names)

shell [project1] project = Project1 type = bacteriome dir_project_ = /home/user/Project1 samples = SAMPLE001,SAMPLE002,SAMPLE003

The database files must be downloaded or compiled, and the custom path must be specified beforehand for the scripts below.

• 08_bacterial_taxonomic_profile.rb
  • db variable
  • Metaphlan database: http://cmprod1.cibio.unitn.it/biobakery4/metaphlan_databases/
• 09_bacterial_contig_taxonomy.rb
  • ppsp_dir variable
  • PhyloPythiaS+ software: https://github.com/algbioi/ppsp/wiki
• 14_pfam_search.split.rb
  • db variable
  • Pfam HMM files: https://www.ebi.ac.uk/interpro/download/Pfam/
• 14_kegg_search.split.rb
  • GHOST-MP database must be compiled beforehand by using KEGG GENES fasta file (e.g. kegg.#{db_date}/genes/fasta/prokaryotes.pep), and be specified to the db path. Accordingly, the db_date parameter in the h_params should be changed to reflect the version of KEGG database used.
• scripts/*
  • The path specification beginning with /home/user must be replaced with the path pointing to the databases (e.g. ko_genes_ = "/home/user/kegg.20180916/genes/ko/ko_genes.list" in assign_kegg.blast.py must be replaced with the path pointing to KEGG database starting with kegg.#{db_date}).
  • Note that the files ko_genes.list, ko_module.list, and ko_pathway.list in the scripts directory is the mapping file from links.tar.gz or genes_ko.list.gz in the database, with the modification that KO identifiers as the first column and linked identifiers as the second.

After the configuration, each step can be run by:

bash ruby 01_trim_qc.HiSeq.rb