OCyRS: Orthologues cyanobacterial gene-associated RNA structure motifs

This repository corresponds to the pipeline of the scientific publication:

"Exploring the regulatory potential of RNA structures in 202 cyanobacterial genomes"
AS Geissler, EC Alvarez, C Anthon, NU Frigaard, J Gorodkin, and SE Seemann
submitted

Purpose

This pipeline download public cyanobacterial genome sequence and annotations from proGenomes. Anchored to orthologues genes, potential adjacently located RNA structure motif are searched for with CMfinder. The pipeline includes additional steps to score the motifs and check for their phylogenetic/biological plausibility with R-scape. As such, the motifs are compared with known bacterial RNA families, the Infernal predicted genomic location of these families, and the respective scores.

Requirements

This pipeline is implemented in a reproducible snakemake workflow, which automatically handles the further software dependencies. The only hard-requirement are

1. A powerful machine or cluster (at least 20 CPU cores 40+ GB RAM)
2. Singularity version $\ge 3.9$
3. Miniconda3 + Mamba
4. Snakemake $\ge 7.9.0$

Linux

Please use either the binary packages coming with your distribution's package system or follow the installation procedures listed by the respective project websites.

MacOS

To install singularity, please consider using a Vagrant box for virtualisation. The installation is straight-forward with Homebrew:

    $ brew install --cask virtualbox vagrant vagrant-manager

With the information in Vagrantfile, singularity can be installed and activated in the current working directory with:

    $ vagrant up
    $ vagrant ssh
    $ cd /vagrant/

Finally, the conda environment must be installed within the virtual box.

    $ wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh
    $ sh Miniconda3-latest-Linux-x86_64.sh

Due to the requirement 'For changes to take effect, close and re-open your current shell.', you must exit and reopen the shell before being able to install mamba.

    $ exit
    $ vagrant ssh
    $ cd /vagrant/
    $ conda install -n base -c conda-forge mamba
    $ mamba install -n base -c bioconda -c conda-forge snakemake

Afterward, the virtual box is fully setup to work with the pipeline.

Dependencies

Download all further software dependencies with:

    $ sh init-singularity.sh

The script downloads the following software images from the Galaxy project software repository:

1. CheckM $1.2.0$
2. CMfinder $0.4.1.18$
3. Infernal $1.1.4$
4. IQ-TREE $2.2.0.3$
5. MUSCLE $3.8.1551$
6. R-Scape $1.4.0$

A container with various R packages renv/renv.sif is pre-build as part of this pipeline. The corresponding conda environment was for maximal reproducibility conda-lock'ed, such that the exact versions are specified for linux x64 computer architectures, in case you would like to re-build the container. An additional container for the TreeShrink software was generated in the folder treeshrinkenv. All additional containers are retrieved via the GitHubs Container Repository of this project.

How to use

Before you start: Please adjust the project_root variable in the config/config.yaml file to your project path.

There are currently two helper script to aid starting snakemake. Use either

1. $ bash run_local.sh for running snakemake on a single powerful machine.
2. $ bash run_slurm.sh to submit jobs to a slurm cluster.

In case of the latter, also consider adjusting config/slurm.yaml to your environment (e.g. partition name). To adapt a similar helper script for alternative cluster solutions, please refer to the snakemake handbook.

Finally, to run the pipeline, first call the download target before initiating all subsequent steps. Together with the container download, the commands to run the pipeline are:

    $ sh init-singularity.sh
    $ bash run_local.sh download
    $ bash run_local.sh all