Reviewed all six original rows. Three direct PMID:22102281 enzyme/process annotations are ACCEPT; three MAT/SAM TreeGrafter claims remain UNDECIDED. Original source fields are untouched. F8U970 maps to AEH41994.1/JF740057 and the fungal UDP-rhamnose epimerase/reductase; the existing OpenAI report researches a different dihydroorotate dehydrogenase. Marked that report WRONG_IDENTIFIER rather than incorporating false mitochondrial/pyrimidine functions. The global Perplexity report for F8U970 (cache70b8d6b9) also researches the wrong target, bacterial PyrR and Synechococcus, after admitting target ambiguity. Its substantive body was read and rejected. Primary evidence carries the decisions.
Read cached Abstract/Introduction/Discussion; the publication cache says full text but omits the Methods/Results. Official force-fetch still yielded incomplete HTML, EuropePMC XML returned HTTP 500, and ordinary PMC open returned a browser challenge. Primary publisher/PMC indexed Results were accessible: https://pmc.ncbi.nlm.nih.gov/articles/PMC3256918/ and https://www.sciencedirect.com/science/article/pii/S0021925820533280. They describe recombinant U4k6dG-ER assays, NADPH-dependent UDP-rhamnose formation and NMR product identification. Exact short excerpt: “NADH substitute for NADPH for the reduction reaction” appears within the explicit negative substitution statement. No rejection of direct IDA is based on incomplete retrieval.
Saved actual tree-source lineage with response hash. Frozen graft PTN008946252 lies in the current plant subtree, below eukaryotic PTN000051875 whose MAT complex/SAM IBD comes from Q9NZL9. Exact F8U970 is absent from the reference tree. The current broad regulator PTN000051877 is not in this graft path. This version/placement distinction is in the neutral focused request; no fabricated target clade or assumed loss is asserted. No NEW annotations added.
Read the complete delivered OpenScientist report and checked the primary human
MAT-complex study PMID:25075345(https://pubmed.ncbi.nlm.nih.gov/25075345/).
The report correctly identifies the sugar enzyme, but recommends removal of all
three MAT-related terms using absence of an assay, relative sequence identity,
and the presence of a different SAM synthetase. These do not exclude a regulatory
second function. Human MAT itself contains separate catalytic and regulatory
subunits. The report's use of mammalian literature does not demonstrate that
all fungi lack any related interaction. Its claimed alignment code is absent
from the delivered HTML/PDF artifacts; the percentages were not adopted.
The primary human structure/solution study reports an alpha4-beta2 assembly and
an essential terminal interaction motif in MATbeta variant V2 (paper numbering
V321/F322/H323). Tail deletion disrupted assembly while retaining secondary
structure. This is more relevant to the transfer than global sequence identity,
but is evidence for that human interface, not a universal diagnostic for fungal
interaction loss. F8U970 has 292 residues and ends QCFERMKKAGVQ: the literal
terminal sequence differs, without claiming homologous residue positions.
The cached entry is older TrEMBL; current entry52 is reviewed Swiss-Prot, as the
report says. Both have sequence version1 and identical sequences. The frozen
source file was not rewritten. Check and hashes are in
the evidence artifact.
All six source objects are preserved. The three direct sugar-enzyme/process rows
remain ACCEPT; three MAT-related rows remain UNDECIDED with revised reasons.
Report incorporation is complete, while the biological and historical-graft
questions remain unresolved. No duplicate research or new annotation was added.