Standard Operating Procedure (SOP)

Title

SOP for Creating New and Updating Existing QC Metric Definitions for WGS-QC Standards v2.0

Version

• SOP Version: 1.0

• Maintained by: GA4GH WGS-QC Workgroup

1. Purpose

The purpose of this SOP is to define a clear, standardized process for GA4GH WGS-QC workgroup members and subject-matter experts to contribute to the development of WGS-QC Standards v2.0. This includes:

  • Creating new QC metric definitions for short-read and long-read WGS

  • Updating or amending existing QC metric definitions from v1.0

  • Differentiating QC metrics by sequencing data type (short-read vs long-read)

This SOP supports the roadmap v2 activities and ensures consistency, traceability, and reviewability of QC metric definitions prior to publication.

2. Scope

This SOP applies to: All contributors to the GA4GH WGS-QC Standards GitHub repository

  • Development of QC metrics for:
    • Short-read WGS
    • Long-read WGS (ONT, PacBio, Constellation)
  • Metric definition work contributing to WGS-QC Standards v2.0

Out of scope:

  • Tool benchmarking or performance evaluation
  • Dataset-specific QC thresholds

3. References

4. Roles and Responsibilities

4.1 Contributors

  • Propose new QC metrics or updates to existing metrics
  • Author metric definitions using the agreed template
  • Submit changes via GitHub pull requests

4.2 Reviewers (WGS-QC Workgroup)

  • Review metric definitions for clarity, correctness, and alignment with consensus
  • Provide feedback and request revisions where required
  • Approve and merge contributions

4.3 Maintainers

  • Manage branches and releases
  • Publish approved metric definitions to gh-pages

5. Overview of Contribution Workflow

  1. Clone the GA4GH WGS-QC repository
  2. Create a new feature branch
  3. Create or update QC metric definition files (.md)
  4. Commit and push changes
  5. Create a pull request (PR)
  6. Review, revise, and merge into develop-definition-v2
  7. Publish approved metrics to gh-pages

6. Metric Definition Requirements (v2.0)

Each QC metric definition must follow the latest workgroup-consensus template and be authored as a Markdown (.md) file.

6.1 Mandatory and Optional Fields

Each metric definition must include the following sections:

  • ID (mandatory): Unique metric identifier (e.g., yield_bp_q30)
  • Description (mandatory): Clear description of what the metric measures
  • Implementation details (mandatory): How the metric is calculated or derived
  • Type (mandatory): Whether the metric value is an integer or float. For float values, the minimal required decimal precision must be defined to ensure consistent reporting across workflows.
  • Functionally equivalent implementation (optional): Alternative implementations/tools
  • Data type (mandatory): short-read | long-read
  • Assembly (mandatory): GRCh37 | GRCh38
  • Version (mandatory): 1.0 | 2.0
  • Sequencing platform / Technology (optional): Illumina | Constellation | ONT | PacBio
  • Associated aligner (optional): Free text or controlled list (e.g., BWA-MEM, DRAGEN, Minimap2..)
  • Associated variant caller (optional): Free text or controlled list (e.g., GATK, DRAGEN, DeepVariant..)

6.2 Versioning Guidance

  • Existing v1.0 metrics that are amended or extended for v2.0 must:
    • Clearly state Version: 2.0
    • Document changes or clarifications in the implementation details
  • New metrics introduced in Roadmap v2 must be labelled as Version: 2.0

7. GitHub Contribution Procedure

7.1 Clone Repository

git clone https://github.com/ga4gh/quality-control-wgs.git
cd quality-control-wgs

7.2 Create Feature Branch

Create a new branch for your contribution:
git checkout -b feature/<short-description>

Example:

git checkout -b feature/long-read-coverage-metric

7.3 Navigate to Metric Definitions Directory

cd docs/metrics_definitions

7.4 Create or Update Metric Definition

  • Create a new .md file using the metric definition template, or
  • Amend/update an existing metric definition to:
    • Correct errors
    • Improve clarity
    • Differentiate short-read vs long-read sequencing File naming should be descriptive and consistent with existing metrics.

8. Commit and Push Changes

Stage and commit your changes:

git add <file(s)>
git commit -m "Add/update QC metric definition for <metric name>"

Push your feature branch to origin:

git push origin feature/<short-description>

9. Pull Request (PR) Process

  1. Create a Pull Request (PR) from your feature branch
  2. Target branch: develop-definition-v2
  3. Provide a clear PR description including:
    • Purpose of the metric
    • Whether it is new or an update to v1.0
    • Any sequencing-type-specific considerations
    • Links to or references for any relevant GitHub issues, pull requests, or discussion threads
## Summary 
Brief description of the QC metric being added or updated. 

## Metric Scope 
- Metric name / ID: 
- New metric or update to existing v1.0 metric: 
- Data type: short-read | long-read 

## Rationale 
Explain why this metric is required or why changes are proposed. 

## Implementation Notes 
Key implementation details, assumptions, or sequencing-technology-specific 
considerations. 

## Related Discussions 
- GitHub issue(s): <link> 
- Pull request(s): <link> 
- Workgroup discussion / meeting notes (if applicable): <link>

Contributors are encouraged to use the following template when creating a pull request:9.2 Review and Approval

  • PRs will be reviewed by WGS-QC workgroup members
  • Contributors may be asked to revise content based on feedback
  • Once approved, PRs will be merged into develop-definition-v2 branch

10. Publication

  • Approved metric definitions will be published via the gh-pages branch
  • Published content represents the official WGS-QC v2.0 product specification

11. Change Management

  • All changes must be tracked via GitHub commits and PRs
  • Major updates or breaking changes should be discussed during workgroup meetings

12. Contact

For questions or clarification, please contact the GA4GH WGS-QC workgroup via the GitHub repository or scheduled workgroup meetings.

List of Potential Long-read QC Metrics

The following non-exhaustive list represents candidate QC metrics for consideration as part of the WGS-QC v2.0 roadmap. These metrics are provided to guide discussion, prioritization, and contribution by the workgroup and domain experts.

Read and Coverage-level Metrics::

Read N50
Yield (bp) 
Yield (reads ≥ N bp): 0, 10,000, 50,000, 100,000 
Read number ?
Reads mapped 
Reads mapped (%) 
Median read quality / Mean read quality 
Median read length / Mean read length 
Autosomal depth (mean), non-gap white listed regions 
Bases with ≥ N-fold coverage (e.g. 5X / 10X / 15X) 
Genome coverage uniformity 
Contamination

Variant-level Metrics::

SNV count 
Indel count (short indels | long indels) 
Heterozygous / Homozygous ratio (SNVs | Indels) 
Transition / Transversion (Ti/Tv) rate