Variant Calling using Strelka

1. Install and Activate Strelka

conda create -n strelka -c bioconda strelka
conda activate strelka

This creates a new Conda environment named strelka and installs Strelka2 from the Bioconda channel.

2. Index the Reference Genome

samtools faidx GCF_018350215.1_P.leo_Ple1_pat1.1_genomic.fna

faidx stands for FASTA indexing. Skip this step if the .fai index file already exists.

3. Run Strelka (Two-Step Process)

Step 1: Configuration

Check if Strelka is installed properly:

which configureStrelkaGermlineWorkflow.py

Configure the workflow for a single sample:

configureStrelkaGermlineWorkflow.py \
  --bam SRR836370_deduplicated.bam \
  --referenceFasta GCF_018350215.1_P.leo_Ple1_pat1.1_genomic.fna \
  --runDir SRR836370_strelka_germline

Explanation:
configureStrelkaGermlineWorkflow.py prepares the pipeline and generates the necessary files.
--bam specifies the input BAM file.
--referenceFasta provides the reference genome.
--runDir specifies the output directory for the configured workflow.

Step 2: Workflow Execution

SRR836370_strelka_germline/runWorkflow.py -m local -j 8

This runs the variant calling process using the previously configured workflow.
-m local runs the workflow on the local machine.
-j 8 uses 8 CPU threads for parallel execution.

4. Variant Calling for Multiple Samples

Strelka can also be configured using multiple BAM files.

configureStrelkaGermlineWorkflow.py \
  --bam SRR836370_deduplicated.bam \
  --bam SRR836361_deduplicated.bam \
  --bam SRR11286181_deduplicated.bam \
  --referenceFasta GCF_018350215.1_P.leo_Ple1_pat1.1_genomic.fna \
  --runDir multi_sample_strelka

Then run the workflow:

multi_sample_strelka/runWorkflow.py -m local -j 8

Note: The output directory will contain the variant calling results generated by Strelka.

Extra Note

The reference genome and other workshop datasets are available from the Wildlife Genomics 4 Zenodo record.

Zenodo dataset: https://zenodo.org/records/21871083