Reconcile wheat crown-rot source-cache discrepancies and experimental timing #1945

Finding

Existing PMID_41504167.txt contains wording absent from freshly retrieved PMC12833866 XML: its abstract says Full-length ITS and 16S profiling revealed..., while the current abstract says Amplicon sequencing revealed...; its bacterial paragraph assigns Paenibacillus/Devosia/Paraburkholderia/Rhizobium enrichment to SMC+Fp, while the current Results assign it to SMC. The cause of the discrepancy is not established; preserve provenance rather than silently overwriting history. The article itself differs on 45 days after pathogen versus SMC inoculation, names both Spearman and SparCC network inference, and has differing alpha-diversity labels. Audit source versions and affected non-graph claims, without treating a cache-based validator pass as fresh primary verification. License is CC BY-NC-ND4.0. Prior1035/1037 address different participant/CI findings.

Evidence And Limits

https://pmc.ncbi.nlm.nih.gov/articles/PMC12833866/

Four whole canonical records reviewed against three unique primary articles. Fresh abstracts and Crossref identities checked for all records; complete wheat C1/C6 and crown-rot main bodies read, including methods, results, discussion, table text and captions. C1/C6 retrievals have identical body hashes. Tomato changes are abstract-bounded: EuropePMC exposed no PMC article and the publisher open returned403. No exhaustive full-text absence claim, original figure-pixel or separate-supplement audit, raw-data reanalysis or new optimization protocol. Local searches include hidden/ignored files with credentials excluded; all-state issue titles/bodies checked.

Lifecycle

Separate non-graph follow-up; remain open until separately verified. Source-based self-adversarial review is not independent approval.