MARCIE (Mapping, vARiant Calling and genotyping pIpEline)

Introduction

This is a simple pipeline written in Bash for performing reads' mapping (Illumina paired-end reads) on a reference genome and subsequent filtering of BAM alignment files, short germline variant calling and combined genotyping of several samples (using GATK). The pipeline has been tested using GATK v. 4.2.6.1 installed within a conda environment (see below).

Specifically the steps are the following:

1. Check for software dependencies
2. Run mapping of all libraries using the BWA mem algorithm in paired-end mode
3. Converting, sorting and filtering of BAM files according to provided threshold for mapping quality (MQ) - each library is processed separately after each mapping is finished
4. Marking duplicates with Picard
5. Short germline variant calling with HaplotypeCaller (GATK toolkit)
6. Combining the GVCF files (CombineGVCFs)
7. Genotyping the combined GVCF files (GenotypeGVCFs)

Requirements

1. BWA mapper
2. GATK toolkit (for variant calling and subsequent genotyping)
3. Sambamba tool for processing and filtering BAM files (sorting, mapping quality filtering)
4. Picard (for marking duplicates)
5. Samtools (for indexing the genome fasta files)

The script requires that all above software dependencies are on the path and performes an initial check for dependencies. The easiest way to run the pipeline is to generate a conda environment, install all dependencies there and run the pipeline while the environment is activated.

#create and activate env
conda create -n mapping_varcall_genotype
conda activate mapping_varcall_genotype
#install requirements
conda install -c bioconda sambamba
conda install -c bioconda samtools
conda install -c bioconda picard
conda install -c bioconda bwa
conda install -c bioconda gatk

Formatting of input files

The script takes as input a list of paired-end Illumina files representing different samples or libraries (e.g., see below sample_1, sample_2) and a genome assembly file in fasta format. A directory path with all the reads' fastq files has to be provided as a command line argument to the bash script. Make sure the samples/libraries have the following naming format in your reads' directory (replace sample_1, sample_2 with your sample codes). The pipeline assumes gzipped fastq files of the Illumina reads. If another format or naming is used the pipeline will fail.

Example 1:

sample_1.R1.fastq.gz
sample_1.R2.fastq.gz
sample_2.R1.fastq.gz
sample_2.R2.fastq.gz

Example 2:

codeA.R1.fastq.gz
codeA.R2.fastq.gz
codeB.R1.fastq.gz
codeB.R2.fastq.gz

Usage

Provide the following inputs to the ´marcie.sh´ script with the following order (the order is important but the variable names might differ. You could also provide the paths to these files as command line arguments to the script without assigning them to variables beforehand):

Example:

#set variables
genome=example_genome.fasta
READS_DIR=/home/user/reads
SPECIES=species_sp
THREADS=10
#threshold for filtering out low quality mappings
MQ=10

#run pipeline
./marcie.sh $genome $READS_DIR $SPECIES $THREADS $MQ

Important: The genome fasta file should have the suffix ".fasta".

The species name variable is important as it will be the prefix of your output files.