This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Any genomics or bioinformatics lab open to reviewing a small research-only DNA traceability pilot?

Hi everyone,

I’m looking for a genomics, molecular biology, or bioinformatics lab/team that may be open to reviewing a small research-only pilot around DNA workflow traceability and reproducibility.

The pilot is intentionally small and non-clinical. It uses a short public or simulated DNA sequence, compares a reference sequence with a sample sequence, records simple SNV differences, logs the operation path, and generates a structured trace report with input/output hashes and run metadata.

I am not looking for clinical validation, patient data, or diagnostic use. I’m looking for technical feedback from a lab or bioinformatics team on whether this type of trace report could be useful for documenting, reviewing, or reproducing genomic workflows.

The ideal feedback would be:

  • what metadata is missing
  • whether the report format is useful
  • what would make it credible in a lab setting
  • whether a small public/simulated dataset pilot would make sense

If anyone here works in or knows a lab/team that might be open to this kind of research-only review, I’d really appreciate guidance or an introduction.

Thanks.

metadata workflow reproducibility genomics variant-analysis
• 373 views • link
updated by manaswwm manaswwm • written by Business • Business

I have worked with genomics datasets, especially SNPs, so I can volunteer to take a look. But I would need more information on what I am looking at? And what are you building it for?

• 0 views • link
written by manaswwm manaswwm

Thank you, I really appreciate that.

I am building ZGL OS, a research-only prototype for genomic workflow traceability and reproducibility. It is not a clinical diagnostic tool and does not replace existing bioinformatics pipelines.

The goal is to package a small genomic workflow into a reviewable evidence object: inputs, outputs, variant records, JSON trace, hashes, operation history, and a short reproducibility report.

For the current case study, I used a public/synthetic TP53-centered 1M-base DNA sequence with 3 controlled SNVs. I would like feedback on:

  1. whether the output/report is understandable from a genomics/bioinformatics perspective;
  2. whether the JSON trace and evidence package contain the right information;
  3. what metadata or validation steps are missing;
  4. whether this kind of traceability layer could be useful for labs or researchers working with SNP/SNV workflows.

I can share a short PDF summary and demo first, then the technical evidence package if you are interested.

• 0 views • link
written by Business • Business

Sure, sounds interesting. I have a couple of busy weeks coming up, and I will try my best to revert with my opinion about this by the end of the first week of June. I can take a look at your PDF summary. I cannot promise if I can run the demo of your software, as I am on a Windows machine at my workplace.

If this works for you, then send me an email at - my exact user name [at] gmail dot com

• 0 views • link
written by manaswwm manaswwm

1 answer

Just post it here and on Reddit and hope that someone will look into it. It's, with all due respect, probably yet another vibe-coded AI slop, so don't expect much feedback.

Ask yourself whether it subsrtantially improves over Cromwell, Nextflow, Snakemake, CWL and all these.

• 0 views • link
written by ATpoint ATpoint

Log in to answer this question.

More posts like this