Wetlab2Variations Workflow Demonstrator

Rare disease researchers’ workflow is that they submit their raw data (fastq), run the mapping and variant calling RD-Connect pipeline and obtain unannotated gvcf files to further submit to the RD-Connect GPAP or analyse on their own.

This demonstrator focuses on the variant calling pipeline. The raw genomic data is processed using the RD-Connect pipeline (Laurie et al., 2016) running on the standards (GA4GH) compliant, interoperable container orchestration platform.

This demonstrator will be aligned with the current implementation study on Development of Architecture for Software Containers at ELIXIR and its use by EXCELERATE use-case communities

For this implementation, different steps are required:

1. Adapt the pipeline to CWL and dockerise elements
2. Align with IS efforts on software containers to package the different components (Nextflow)
3. Submit trio of Illumina NA12878 Platinum Genome or Exome to the GA4GH platform cloud (by Aspera or ftp server)
4. Run the RD-Connect pipeline on the container platform
5. Return corresponding gvcf files
6. OPTIONAL: annotate and update to RD-Connect playground instance

N.B: The demonstrator might have some manual steps, which will not be in production.

RD-Connect pipeline

Detailed information about the RD-Connect pipeline can be found in Laurie et al., 2016

The applications

1. Name of the application: Adaptor removal
   Function: remove sequencing adaptors
   Container (readiness status, location, version): cutadapt (v.1.18)
   Required resources in cores and RAM: current container size 169MB
   Input data (amount, format, directory..): raw fastq
   Output data: paired fastq without adaptors

2. Name of the application: Mapping and bam sorting
   Function: align data to reference genome
   Container : bwa-mem (v.0.7.17) / Sambamba (v. 0.6.8 )(or samtools)
   Resources :current container size 111MB / 32MB
   Input data: paired fastq without adaptors
   Output data: sorted bam

3. Name of the application: MarkDuplicates
   Function: Mark (and remove) duplicates
   Container: Picard (v.2.18.25) Resources: current container size 261MB
   Input data:sorted bam
   Output data: Sorted bam with marked (or removed) duplicates

4. Name of the application: Base quality recalibration (BQSR)
   Function: Base quality recalibration
   Container: GATK (v.3.6-0) Resources: current container size 270MB
   Input data: Sorted bam with marked (or removed) duplicates
   Output data: Sorted bam with marked duplicates & base quality recalculated

5. Name of the application: Variant calling
   Function: variant calling
   Container: GATK (v.3.6-0) Resources: current container size 270MB
   Input data:Sorted bam with marked duplicates & base quality recalculated
   Output data: unannotated gvcf per sample

6. (OPTIONAL)Name of the application: Quality of the fastq
   Function: report on the sequencing quality
   Container: fastqc 0.11.8 Resources: current container size 173MB
   Input data: raw fastq
   Output data: QC report

Licensing

GATK declares that archived packages are made available for free to academic researchers under a limited license for non-commercial use. If you need to use one of these packages for commercial use. https://software.broadinstitute.org/gatk/download/archive