agentsclimarketplace

Workflow output validation and qa

Skill HolobiomicsLab/asb-skill-collections/packs/metabolomics/lc-ms/skills/workflow-output-validation-and-qa

Curated, evidence-grounded skill and software-tool collections for scientific AI agents, generated by the AgenticScienceBuilder

Install
npx -y skills add HolobiomicsLab/asb-skill-collections --skill workflow-output-validation-and-qa

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 14 stars14 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Use when after executing a Nextflow-based MS-DIAL workflow on .mzML LC-HRMS metabolomics data using Docker or Singularity container backends.

The file declares its own license as CC-BY-4.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

7.8 KB, ~1.4k tokens by cl100k_base, as published. Nobody here has run it

workflow-output-validation-and-qa

Summary

Verify successful execution of containerized LC-HRMS metabolomics workflows by inspecting presence, format, and integrity of intermediate and final outputs (MS-DIAL .msdial → .tsv files and MSFLO results). This skill ensures reproducibility and detects failures before downstream analysis.

When to use

After executing a Nextflow-based MS-DIAL workflow on .mzML LC-HRMS metabolomics data using Docker or Singularity container backends. Use this skill when you need to confirm that peak detection, chromatogram alignment, and metabolite identification steps completed without silent failures, and to validate that output files are accessible and in expected formats before loading into spreadsheet or statistical software.

When NOT to use

  • Workflow execution terminated with explicit error messages before completion — validation assumes the workflow was allowed to finish; inspect error.log instead.
  • Results directory is empty or inaccessible — this indicates a container initialization or file system permission problem, not an output validation issue.
  • You are validating raw .mzML files before processing — use input data quality checks, not output validation.

Inputs

  • .mzML LC-HRMS raw mass spectrometry data files
  • MS-DIAL configuration file (msdial_params.txt)
  • MS-FLO configuration file (msflo_params.ini)
  • MS1 and MS2 spectral libraries (ms1_lib.txt, ms2_lib.msp)

Outputs

  • MS-DIAL .msdial output files (converted to .tsv format)
  • MSFLO results files
  • execution_report.html (Nextflow runtime and resource summary)
  • execution_timeline.html (per-process execution timeline)
  • execution.log (detailed workflow metadata and process-level logs)

How to apply

After workflow completion, navigate to the designated results directory and verify the presence of expected MS-DIAL output files (.msdial, converted to .tsv for spreadsheet compatibility) and MSFLO output files. Check that all output files are non-empty and readable. Cross-reference the execution log (execution.log or execution_report.html) to confirm that all expected processes completed without error. Examine the execution_timeline.html to ensure each process stage (peak detection, alignment, identification) executed in correct order. If using a high-performance computing environment like HiPerGator with Singularity, verify that container initialization and resource allocation messages appear in execution logs. Validate output file schemas by opening a sample .tsv output in a text editor or spreadsheet application to confirm it contains expected metabolite feature columns and intensity values.

Related tools

  • Nextflow (Workflow orchestration and process execution tracking; produces execution_report.html, execution_timeline.html, and execution.log artifacts for validation) — https://www.nextflow.io/
  • MS-DIAL (Peak detection, chromatogram alignment, and metabolite identification; produces primary .msdial output files validated in results directory)
  • MSFLO (Post-processing and statistical analysis of MS-DIAL features; produces secondary results files included in validation check)
  • Docker (Container backend for local execution; validation confirms container initialization and file mounts in execution logs) — https://docs.docker.com/engine/installation/
  • Singularity (Container backend for HPC environments; validation confirms Singularity container binding and resource allocation in execution logs) — https://www.sylabs.io/guides/3.0/user-guide/

Examples

nextflow run main.nf -profile docker > logs/execution.log && ls -lah results/*.tsv && head -n 5 results/sample_output.tsv

Evaluation signals

  • Presence of .tsv output files in results directory matching expected count and sample naming convention from input data
  • Execution logs report 'Completed successfully' or equivalent for all MS-DIAL and MSFLO processes; no 'FAILED' or 'ERROR' entries
  • execution_report.html shows all processes in green/completed state and total execution time > 0
  • Sample .tsv file contains metabolite feature columns (RT, m/z, intensity) with numeric values and non-zero row count
  • execution_timeline.html shows processes executed in logical sequence: peak detection → alignment → identification → MSFLO

Limitations

  • Workflow does not validate correctness of results against ground truth or reference standards — only confirms file presence and format. Output values may be scientifically invalid.
  • No explicit validation of MS1/MS2 library matches or metabolite identification accuracy is performed; inspect MSFLO confidence scores manually.
  • File name conventions are strict; special characters in input file names cause silent process skipping and false validation passes. Use underscores only.
  • Initial release (Nextflow4MS-DIAL v1.0) lacks explicit reproducibility testing documentation; validation is heuristic-based on log inspection, not formal test suite.
  • Execution timeline and resource reports vary between Docker (local) and Singularity (HPC); HPC-specific resource allocations must be verified against job scheduler output (Slurm) separately.

Evidence

  • [other] Verify workflow completion and validate presence of MS-DIAL and MSFLO output files in the designated results directory.: "Verify workflow completion and validate presence of MS-DIAL and MSFLO output files in the designated results directory."
  • [readme] Example outputs are stored in the results folder. The file extensions for produced .msdial files have been changed to .tsv so the files can be opened in spreadsheet software.: "Example outputs are stored in the results folder. The file extensions for produced .msdial files have been changed to .tsv so the files can be opened in spreadsheet software such as Microsoft"
  • [readme] execution_report.html summarizes workflow runtime and computational resource usage; execution_timeline.html shows the execution timeline for each process; execution.log includes metadata such as workflow version, parameter settings, resource allocation, container information, and process-level logs.: "execution_report.html summarizes workflow runtime and computational resource usage. execution_timeline.html shows the execution timeline for each process. logs/execution.log is an example log"
  • [readme] To avoid unexpected errors, do not use special characters in file names. Underscores are safe to use.: "To avoid unexpected errors, do not use special characters in file names. Underscores are safe to use."
  • [other] containerized workflow MS-DIAL → MSFLO enables processing of .mzML LC-MS metabolomics data: "Enabling processing .mzML LC-MS metabolomics data through a containerized MS-DIAL workflow."

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 327,132. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.