Feedback from PHA4GE data structures WG

[cbarcl01]

Hi QC WGS team!

Thanks for creating this resource Members for the PHA4GE Data Structures working group reviewed this product and collected feedback to represent the community.

A document detailing feedback available here, but in summary:

• Metrics scope: Could standard more clearly distinguish between:
  • QC methods/analysis outputs (e.g., Q30 score, coverage depth).
  • QC outcomes (e.g., pass/fail, acceptable/insufficient).
• Standard has a human-centric bias: Current fields (e.g., autosomal genotype metrics) are not broadly applicable to microbial/pathogen genomics; structure needs to be more flexible.
• Pipeline sensitivity & auditability: Add mechanisms to record the specific tools, methods, and software versions used to generate QC metrics?

Implementation feedback:

• Improve documentation clarity: single summary table of all metrics in the README would be really helpful.
• Provide practical guidance: definitions with edge cases, pitfalls, and recommended sources.
• Deliver validation support: JSON Schema or equivalent tooling plus ready-to-use scripts/checkers could be useful to the community (future dev).

Thanks!
PHA4GE DSWG

[justinjj24]

Metrics scope:

We currently provide both the QC metric definitions and the implementation details in the documentation for each metric, such as Q30 and mean coverage. https://ga4gh.github.io/quality-control-wgs/metrics_definitions/
We define the standards and provide a standalone QC workflow that satisfies the GA4GH quality control metrics. However, we do not intend to determine or report whether the data is “good” or “bad" / "pass" or "fail".

The workgroup’s current focus is on the human genome. While there are no immediate plans to extend the standard to other species, future versions could consider broader applicability if there is community interest.

Yes, the methods, workflows, tools, software are recorded in the README.md and container conda_env.yml used to generate QC metrics are documented in our version controlled implementation Git repository. We could explore ways to make this information even more accessible or user-friendly in future updates.

Implementation feedback:

Both the standard and the implementation Git product specification pages already list all the metrics. We could also add a summary table in the README.md (covering short-read post-alignment and post-variant calling QC metrics) for improved clarity.
https://ga4gh.github.io/quality-control-wgs/metrics_definitions/
https://c-big.github.io/NPM-sample-qc/metrics.html

Yes, we will include a schema checker/validator as part of the QC metric report JSON.

Provide practical guidance: definitions with edge cases, pitfalls, and recommended sources.
see #108 (comment)

[cbarcl01]

Thanks for the reply to the feedback, I'll update the PHA4GE community. In the meantime feel free to close this ticket :)