Skip to main content

VAR Pipeline Guide

This guide provides step-by-step instructions for setting up and running the UCGD VAR pipeline. Each input parameter for Nextflow execution is explained in detail.

Prerequisites​

  • Access to AWS S3 buckets.
  • Access to the Mosaic server.
  • Access to reference FASTA and BED files.
  • Access to the EC2 environment with necessary permissions.

Step-by-Step Instructions​

If you decide not to use the given variant-caller.sh script here are explanation of each prompt:

1. Start the Script​

Run the script to initiate the setup.

2. Provide Required Inputs​

You will be prompted to enter the following inputs interactively:

a. Source File Path​

  • Prompt: Enter path to source file (e.g. s3://...):
  • Description: Full path to the source file for input.
  • Example: s3://your-bucket/source_files_ids.txt

b. Complete Path​

  • Prompt: Enter complete path (e.g. s3://...):
  • Description: S3 path where final output files will be stored.
  • Example: s3://your-bucket/

c. Mosaic Project ID​

  • Prompt: Enter Mosaic Project ID:
  • Description: Unique identifier for the Mosaic project.
  • Example: 1685

d. Project Name​

  • Prompt: Enter Project Name:
  • Description: Human-readable project name.
  • Example: M1685-Project

e. Mosaic Config Server​

  • Prompt: Enter Mosaic config server:
  • Description: Mosaic server configuration URL or IP.
  • Example: ucgd-mosaic

f. Reference Masked FASTA​

  • Prompt: Enter reference masked FASTA path (alignment step):
  • Description: Path to the masked reference FASTA for alignment.
  • Example: /common/Reference/[build]/*_MASKED.fasta

g. Reference FASTA​

  • Prompt: Enter reference FASTA path (CRAM generation step):
  • Description: Path to the full reference FASTA used for CRAM generation.
  • Example: /common/Reference/[build]/*_.fasta

h. Align Seed Value​

  • Prompt: Enter align seed value (e.g. 19):
  • Description: Align seed value for alignment step.
  • Example: 19

i. PCR Library Type​

  • Prompt: Enter PCR library type (e.g. pcr/nopcr):
  • Description: Specify the type of PCR library.
  • Example: nopcr

j. Glnexus YAML Path​

  • Prompt: Enter Glnexus YAML path (Deepvariant yaml):
  • Description: Path to Glnexus configuration YAML file.
  • Example: /common/Deepvariant/

k. Region BED File​

  • Prompt: Enter region BED file path:
  • Description: Path to the BED file defining regions of interest.
  • Example: /common/Regions/[build]

Directory Setup​

  1. Work Directory

    • Path: /common/nextflow/$MOSAIC_PROJECT_ID
    • Created and permissioned the needed EFS directory for Nextflow execution.

  2. Processing Directory

    • Named: [M]$MOSAIC_PROJECT_ID
    • Contains the linked UCGD-Cloud codebase.

Nextflow Command Construction​

The following required parameters are assembled into a Nextflow command, but can be used to run an ad hoc process as needed:

  • --source_file: Source input file path.
  • --complete: Path for final outputs.
  • --mosaic_project_id: Mosaic project identifier.
  • --mosaic_server: Mosaic server config file name.
  • --project: Project name.
  • --reference_masked: Masked reference FASTA.
  • --align_seed: Alignment seed value.
  • --pcrLibrary: PCR library type.
  • --reference: Full reference FASTA.
  • --glnexus_yaml: Glnexus configuration file.
  • --regionBeds: BED file for regions.
  • -w: Work directory.
  • -profile cloud: Execute using cloud profile.
  • -resume: Resume from previous runs.

Execution Confirmation​

  • Review the generated Nextflow command.
  • Confirm execution by typing yes.

If confirmed:

  • Logs into Docker.
  • Executes the Nextflow pipeline.
  • Monitors execution and reports success or failure.

If not confirmed:

  • Cancels the command execution.

Important Notes​

  • File Validation: Ensure all paths exist and are accessible.
  • Permissions: Ensure necessary AWS and EC2 permissions.
  • Consistency: Validate consistency of reference files and project setup.
  • EFS: After process completion, you must remember to delete your /common/nextflow/M[project] directory, or you will incure AWS EFS charges.