makrez/meta-genomisc-snakemake

Analysis Pipeline for Metagenomic assembly

Pipeline Description

This Pipeline performs metagenomics assembly based on pooled reads from several samples.

It performs the follwing steps:

1. QC: Software trimmomatic. Adapterclip for Nextera Libraries.

2. Host removal with bowtie2

Here, two different

3. Metaassembly of the genomes with metaSPAdes.

Files

Config file

This file is in JSON format.

Tool specifications

• {tool}_version: the specific version used for the tool

• {tool}_threads: the number of threads that are used

• {tool}_hours: the maximum number of hours before the job gets killed

• {tool}_mem_mb: the maximum number of RAM (in megabytes) to be used by this tool

• short_sh_commands_threads: Number of threads used for short bash commands. Usually used when one-liners are called.

• short_sh_commands_hours: Number of hours used for short bash commands.

Data specifications

• DataFolder: Folder where the data is located

• extension: File extension (only tested with .gz files, however, it could be fastq.gz or anything.gz)

• Lanes: Number of lanes that the samples were sequenced. # //TODO: This may produce errors if e.g. sample_1_L1_[...] and sample_2_L2_[...] are present, but not sample_1_L2_[...] and/or sample_2_L1_[...]

• mates: Paired-end identifier (e.g. R1, R2)

• defaultIlluminaExt: "001" //TODO: If this is not present, it will produce " __ " and throw errors

Reference Files and other Resources

• Host reference: The location of the file reference_list.txt (see Reference_files and resources)

Test parameters

These parameters are used for all tools for testing purposes with a small sample data set.

• testing_threads: Number of threads used for ALL tools in the testing case

• testing_hours: Number of hours used for ALL tools in the testing case

• testing_mem_mb: Number of memory in MB used for ALL tools in the testing case

Reference files

The reference files have to be added manually to the pipeline.

The fasta files have to be put into the folder analysis/02_filter_host/reference and a text file has to be generated. The structure of the .txt file looks as follows:

ref1.fasta,ref2.fasta # add as many as needed

Once the fasta files are in the folder, the easiest way to create reference_list.txt is to run the bash command:

ls | grep 'fasta' | paste -sd ',' > reference_list.txt

Creating the test environment

The following commands should produce a test folder. Inside this folder, the same logic is reproduced but the data folder and the configuration about threads, memory and running time are changed.

chmod + x sync_test.sh # requires rsync
./sync_test.sh

This test can be run on the IBU cluster with the command:

cd test;
module load Utils/snakemake
/opt/cluster/software/Conda/miniconda/3/bin/snakemake --printshellcmds --drmaa " --partition=phshort --ntasks=1 --mem={resources.mem_mb} --cpus-per-task={threads} --time={resources.hours}:0 --mail-type=END,FAIL " --latency-wait 300 --jobs 1 --jobname <jobname>_{jobid}

If this test finishes successfully, the whole script can be run.