R2A graph: remove MOPS-misattributed cross-feeding and unsupported antagonism-rescue arrows #1532

KBase_Synthetic_Bacterial_Community_R2A.yaml, CommunityMech:000039: all three nodes and both arrows reviewed.

The R2A model contains Pantoea YR343, Pseudomonas GM17 and Sphingobium AP49; Variovorax CF313 belongs to the four-member MOPS model. The CF313 cross-feeding node and both incoming arrows transplant MOPS predictions into R2A. Pairwise agar inhibition by GM17 is directly observed, but beneficial metabolite compensation is only one explanation of liquid coexistence; the primary source explicitly gives liquid-versus-static-agar conditions as an alternative. FBA predicts exchanges and conditional model growth, while metaproteomics detects pathway representation, not transferred flux.

Acceptance: remove the misplaced CF313 node and both arrows; retain the existing Pantoea R2A prediction and exact GM17-to-CF313 R2A-agar assay with source-bounded wording; distinguish model assumptions, medium and assay; preserve canonical membership pending separate correction. Use exact OPEN discussion anchors, append-only history, focused tests, regenerated page and hash-bound outputs/review records. Do not add new causal structure or claim independent approval. Keep this graph issue OPEN until protected-main merge.

Primary: https://pmc.ncbi.nlm.nih.gov/articles/PMC8079826/ (PMID33995895). Whole record and full extracted primary main text, Methods2.1-2.6, Results3.1-3.5, Tables1-2 and main captions read. No original figure pixels, supplements, raw-data or KBase execution audit. Canonical taxonomy must not be guessed from the model schematic.

Dedup:834 all-state issue titles/bodies searched;85 ignored/hidden-inclusive local matching files, excluding docs/site and separately inspected reference caches. Relevant #1527 (different KBase records), #1471 (different CF313 experiment), #650 and #653 inspected. Not every matching local file, historical comment or PR was read. Self-adversarial review is not independent approval.