THOR: nutrient level is the setting, and one record refused (#183, #543) - #544
Open
realmarcin wants to merge 1 commit into
Open
THOR: nutrient level is the setting, and one record refused (#183, #543)#544realmarcin wants to merge 1 commit into
realmarcin wants to merge 1 commit into
Conversation
THOR_Rhizosphere_Model_Community: liquid culture in half-strength tryptic soy broth at 28 °C with vigorous shaking, strains propagated on 1/10-strength TSA, interactions assayed across LB, half-strength TSB and soybean root exudate. The dilutions are in the record because they are chosen, not incidental. The paper's finding is that antagonism among the three members is medium-dependent and appears predominantly under the lower-nutrient condition, so a setup saying "TSB" would drop the half-strength — which is where the interactions are. No system_type or working_volume: the source says "liquid culture" and names the broth, never the vessel or the fill, and the assays run on plates as well as in liquid, so one vessel value would misdescribe half the experiments. Also filed #543. Avena_Rhizosphere_CrossKingdom_SIP_Community ranked top of the candidate list — 95k of full text, 100% member coverage — and must not be curated at all: it is stable-isotope probing of native rhizosphere soil, and the community was never grown in a vessel. No field in the record distinguishes that case. community_origin fails (three NATURAL records are legitimately cultivated, all verified), ecological_state fails (two of them are STABLE), and community_category fails (a RHIZOSPHERE record already carries a setup). The check has to be a human reading the source, which is worth knowing before anyone batches the remaining 13 full-text candidates. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Network integrity findingsWarnings only — a member with no interaction yet, or a participant matched by ontology id rather than by name, or one on a community-level interaction that resolves to no member. Reported, but does not fail the build. The full report is attached to the workflow run as an artifact. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The record curated
THOR_Rhizosphere_Model_Community— liquid culture in half-strength tryptic soy broth at 28 °C with vigorous shaking, strains propagated on 1/10-strength TSA, interactions assayed across LB, half-strength TSB, and soybean root exudate.The dilutions are the point. They are chosen, not incidental: the paper's finding is that antagonism among the three members is medium-dependent and appears predominantly under the lower-nutrient condition. A setup recording "TSB" would drop the half-strength — which is exactly where the interactions are.
No
system_typeorworking_volume: the source says "liquid culture" and names the broth, never the vessel or the fill, and the assays run on plates as well as in liquid, so a single vessel value would misdescribe half the experiments rather than approximate them.The record refused, and why it matters more (#543)
Avena_Rhizosphere_CrossKingdom_SIP_Communityranked top of the candidate list — 95k of cached full text, 100% member coverage, methods keywords throughout.It must not be curated at all. It is stable-isotope probing of native rhizosphere soil: Avena fatua was grown under ¹³CO₂ and soil was sampled at 6 and 9 weeks. The plant had a growth setup; the microbial community was never grown in a vessel by anyone.
No field in the record distinguishes that case. I checked all three before assuming any worked:
community_originNATURALrecords legitimately carry a setup —GLBRC_UFMP(in-house CSTR),Phormidium_Alkaline(photobioreactors),Thiocyanate_Afipia(five CSTRs) — all verified as genuinely cultivatedecological_stateSTABLE, notENGINEERED; filtering onENGINEEREDdrops legitimate recordscommunity_categoryRHIZOSPHERErecord already carries one (Lunar_Martian_Simulant_PGPB_Lettuce_SynCom, a stirred-tank bioreactor)So "was grown in a vessel" versus "was sampled from the world" is not derivable from the KB. The check has to be a human reading the source — worth knowing before anyone batches the 13 remaining full-text ENGINEERED candidates.
I deliberately did not add a test. A guard would need an allow-list naming each legitimate case individually, which is a list of decisions rather than a rule, and would give false confidence to exactly the batch run it is meant to protect.
Checks
just validate— no issuesjust validate-references— 0 issues; both snippets verbatimjust validate-strict,just lint,just check-docs-current— exit 0uv run pytest tests/— 2375 passed, 16 skippedPart of #183. Filed #543.
🤖 Generated with Claude Code