Mapping-by-sequencing

• Installation
• Set up working environment
  • Install conda environment
  • Modules to load on Rackham
• Set up a run
  • Reference genome and snpEff database
  • Parental (control) dataset and mutated datasets
• Run!

This pipeline is intended to emulate artMAP and it's been built by following the protocol described in the artMAP paper (please refer specifically to https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6560221/bin/PLD3-3-e00146-s002.pdf).

Installation

Download this repository to your machine

git clone https://github.com/domenico-simone/mapping-by-sequencing.git 

Set up working environment

This pipeline relies on a bunch of dependencies, namely:

• snakemake
• FastQC
• trimmomatic
• bwa
• samtools
• bcftools
• bedtools
• snpEff

if you already have these tools installed on your system, please skip to the section Set up a run. If you don't, you can choose one of the following options:

• install them on your own
• install the conda environment provided in the mapping-by-sequencing repository (please follow instructions here)
• if you are working on the UPPMAX/Rackham cluster, all these tools can be loaded as modules (please follow instructions here)

Install conda environment

cd mapping-by-sequencing

conda env create -n mbs -f envs/mbs.yaml

This will create a conda environment named mbs which has to be activated every time you want to use the pipeline with one of this commands (depending on your conda installation):

conda activate mbs

# if the above fails, use this:
source activate mbs

When you're done with the pipeline, you may want to deactivate the conda environment with:

conda deactivate

Modules to load on Rackham

These are needed to run the workflow down to SNP calling

module load bioinfo-tools
module load snakemake
module load FastQC
module load trimmomatic
module load bwa
module load samtools
module load bcftools
module load BEDTools
module load snpEff
# others

Set up a run

Reference genome and snpEff database

The reference genome in fasta format should be placed in the folder data/reference_genomes and its file name has to be detailed in the config.yaml file in the ref_genome field.

The snpEff db matching the reference genome should be already installed in snpEff and its name should be detailed in the config.yaml file in the snpEff_db field.

Example: if your reference genome is included in a file My_ref_genome.fa and your snpEff database is called Genus_species, the config.yaml file should look like:

results:    "results"
map_dir:    "map"
log_dir:    "logs"
tmp_dir:    "/tmp"
workdir:    "test"
ref_genome: "My_ref_genome.fa"
snpEff_db:  "Genus_species"

read_processing:
    trimmomatic:
        options: "-phred33"
        processing_options: "LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:100"
        java_cmd: "java"
        java_vm_mem: "4G"
        threads: 4

Parental (control) dataset and mutated datasets

File data/datasets.tab should be filled with data about your samples. Its structure is:

sample  sample_type  library  R1              R2
D1K     control      1        D1K_L1_1.fq.gz  D1K_L1_2.fq.gz
D2K     mutated      1        D2K_L2_1.fq.gz  D2K_L2_2.fq.gz
D2K     mutated      2        D2K_L3_1.fq.gz  D2K_L3_2.fq.gz
D3K     mutated      1        D3K_L2_1.fq.gz  D3K_L2_2.fq.gz
D4K     mutated      1        D4K_L1_1.fq.gz  D4K_L1_1.fq.gz
D5K     mutated      1        D5K_L2_1.fq.gz  D5K_L2_2.fq.gz

where

• sample is the sample name
• sample_type indicates whether the sample is a control (the parental line) or a mutated line
• library takes into account the fact that a sample can have multiple read datasets (libraries). Eg, in the example above, sample D2K has two libraries.
• R1 and R2 are the name of the R1 and R2 file for each library, assuming they are located in the directory data/reads.

Run!

snakemake -rp -j 10

Please refer to the official snakemake tutorial for further examples on how to run a snakemake workflow.