Skip to content

TCM gout compound triage

Verdict

COMP-013 is invalidated and non-runnable. It does not support a compound rank, GUT-LUMINAL VIABLE or other viability label, nominal gut or plasma exposure, target occupancy, percent inhibition, clinical-risk tier, dose conclusion, delivery route, or advancement decision.

The model inherited COMP-004's invalid nominal-concentration/drug-substrate-IC50 shortcut, adapted COMP-007's invalid composite score, failed to preserve target-effect polarity, and mixed direct assays, expression changes, whole-animal phenotypes, and off-target signals as if they were comparable.

The hash-bound tombstone preserves the exact retirement scope. Git contains the retired code, inputs, outputs, and reviews.

What survives

The former input names remain an unranked historical search list:

  • luteolin
  • astilbin
  • rhein
  • emodin
  • aucubin
  • berberine
  • cylindrin
  • chlorogenic acid
  • atractylenolide I

The names are not evidence. Each must be rebuilt from the primary record with exact material identity, source, target or endpoint, effect polarity, assay system, evidence level, exposure status, and mechanism-matched function. Sparse database coverage is a reason to search more broadly, not a negative biological verdict.

The formula-level Si Miao San connection also survives as a research lead. A formula signal does not establish component, target, or synergy attribution.

Current evidence and replacement

TCM-derived gout leads owns the current source-verified evidence boundaries, sourcing and delivery constraints, formula conjecture, and discriminating experiments.

COMP-049 is the pre-run replacement design. Its fixed correction set contains four primary animal studies and one secondary systematic review covering Smilax glabra total flavonoids, emodin, coix seed oil, Plantaginis Semen extract, and modified Simiao decoctions. The planned computation preserves source tier, tested-material identity, polarity, assay context, simultaneous evidence gaps, and deferred work without ranking the records.

COMP-049 has not run. No result should be inferred from its design or inputs.

Falsification boundary

An incomplete evidence record does not show that a material is ineffective. It shows which attribution, exposure, functional, or trial-level evidence is missing. A negative experiment kills only the exact material, exposure, endpoint, and decision rule it tested.