Skip to content

CAAIL field-gap analysis: media-optimization / transcriptomics design (9 candidate additions + 4 corrections) #61

Description

@benjibromberg

Context

These candidate additions and corrections surfaced while designing the media-optimization bundle (cell-ag-media-bundle v0.1, transcriptomics-first media design). They are coverage gaps and small fixes found while building a consolidated data catalog and validation plan against CAAIL. Filing as a field-gap analysis to action later, separately from that design work.

Process (same as the prior gap-analysis issues, e.g. #58, #59): work on a feature branch, route every entry through the CAAIL reviewer agents (citation / claim / classification), verify each claim against the version of record, and hold PR/merge for maintainer go-ahead. The items below are candidates to verify during the work, not asserted facts.

Candidate additions

Cultivated-seafood / mackerel (currently a CAAIL species blank)

  • Mack1 line establishment, Saad et al. 2023, Sci Rep (DOI 10.1038/s41598-023-31822-2; lipidomics on OSF, project T742A). CC-BY.
  • Lim et al. 2024 mackerel serum-free media screen (DOI 10.1021/acssuschemeng.4c03345). NOTE: rights-reserved against AI/TDM; human-mediated reference only.

Metabolomics / ID-mapping infrastructure

  • RefMet (metabolite nomenclature reference).
  • BridgeDb (cross-database identifier mapping).

Spent-media / process datasets

  • Robitaille et al. 2015, CHO spent-media metabolomics (a demo-pipeline stand-in; CC-BY).
  • Narayanan et al. 2025 dataset (Zenodo): a component-to-outcome reference for a future Bayesian-optimization loop.

Methods / software

  • Context-specific GEM extraction methods: GIMME, iMAT, tINIT, RIPTiDe (transcriptome to context-specific model; a documented CAAIL gap).
  • BioContextAI MCP servers (biomedical context tooling).
  • BoTorch / Ax (Bayesian optimization; for the v0.2 search loop).

Candidate corrections (verify against source first)

  • Pig.md: reconcile the pcPigGEM2025 vs pcPigMNet2025 naming for the proteome-constrained pig GEM.
  • Chicken.md: Abasht metabolite identifiers are KEGG / HMDB / CAS, not PubChem.
  • CrossSpecies.md: the Combe hydrolysate NMR dataset license is CC-BY-NC-4.0.
  • Confirm and note that the 10.64898 bioRxiv DOI prefix is legitimate (it is the prefix used by the "Real Science Is Harder Than Benchmarks" entry and the Venkatarama DMFA paper), so it should not be flagged as atypical.

Explicitly out of scope

  • BDFAT (bovine de-differentiated fat RNA-seq): permission-gated; do not add without owner permission.
  • CriBSC RNA-seq: excluded as bad data.

Metadata

Metadata

Assignees

No one assigned

    Labels

    documentationImprovements or additions to documentationenhancementNew feature or request

    Type

    No type

    Fields

    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions