Completed all 35 seeded rows. Full PMID:20424176/PMID:23221833 establish native sulfite reduction, sulfur flux and sulfite-toxicity protection; reduced-dye assays are not native donor-pairing assays. Full PMID:27189994 Figure 3 directly shows AtSiR-YFP predominantly stromal without preferential DAPI colocalization; the Phaseolus CEP fusion redirects it to nucleoids. Remove the two nucleoid transfers. DNA binding remains UND becausePMID:17371503 demonstrates maize DNA affinity despite nonnucleoid localization; repression by pea DNA compactionPMID:11997391 is not established for Arabidopsis and is overannotated. The sulfide-oxidation BP is opposite to demonstrated physiological chemistry (live GO definition checked). Inspected full PMID:28887381 and all seven deposited XLSX files: SIR occurs in Table S1 row 1215, localization plastid, no True membrane/Dual Localized flag. Plasma membrane is over-specific; reproducible extraction is committed in SIR-source-analysis.PMID:20018591 full text was normally force-fetched; target-specific copper supplement unavailable, so copper remains UND, not dismissed because iron is the catalytic metal. Cytosol, envelope, apoplast, cold response and stromules remain unresolved where the accessible source does not provide target-level evidence.PMID:16923014/PMID:21166475/PMID:12938931/PMID:18538804 remain abstract-only; no wrong-gene or contamination conclusion is inferred.
Research provenance: parent already attempted Falcon (403) and Perplexity-lite (401 quota) for this batch; no generated report was fabricated. Existing PSEPK OpenScientist reports were used only as discovery. Source caches were fetched normally and never hand-edited. All original annotation payloads were preserved, with a retirement marker only for the existing CysI donor-specific NEW proposal.
Primary comparative donor corroboration: PMID:26920048 reports maize SiR–Fd structures and interface-mutant effects on Fd-dependent turnover/binding. This supports the plant donor class but is explicitly not an Arabidopsis partner-pairing experiment.
Access follow-up: normal publisher URL fetches for cold/stromule, cytosolic-proteome and chloroplast-envelope papers returned HTTP403; the apoplast DOI route yielded only an empty redirect shell. These are not full-text evidence and are excluded from the source manifest. Copper full text was recovered by the normal PMID fetcher, but the supplementary-file endpoint returned no target table.
Corrected the cytosol NOT rationale: the claim excludes cytosol, and stromal imaging alone cannot establish that exclusion. It remains UNDECIDED. Each source finding now states what its own quote establishes. Added a structured residual DNA-binding gap, distinguishing maize/pea assays from Arabidopsis stromal imaging, with a proposed direct assay.
All original annotation payloads and row order are preserved. Focused validation and automatic record-section backfill pass; generated pages and append-only history record this follow-up.