Augmented the existing review without replacing its 50 source annotation identities or historical GOA. The previous review had 27 ACCEPT, 9 KEEP_AS_NON_CORE, 3 MODIFY and 11 REMOVE actions. The current review has 14 ACCEPT, 8 KEEP_AS_NON_CORE, 21 MODIFY and 7 UNDECIDED actions, with no NEW rows. Source-specific evidence and access limits are in DHCR24-evidence-guide.md.
The enzyme evidence supports one delta-24 sterol reductase/cholesterol-synthesis unit at the ER membrane. Current GO distinguishes the delta-24(28) reaction from the demonstrated human delta-24 activity. The former source IDs remain unchanged, with specific replacements. The separate p53-binding unit is supported by the original study's primary abstract and genuine publisher supplement, including human recombinant constructs and separation-of-function experiments. DHCR7 association is retained as enzyme binding; its unresolved cellular regulatory mechanism is not a duplicate catalytic core.
The previously removed PGRMC1 pairs are supported by an exact curated crosslink record and the OpenCell author interaction dataset. No justified narrower partner class was found, so they are kept as non-core under the explicit ClinGen project instruction. That scoped instruction supersedes generic-only removal advice. These two rows may retain ordinary validator advisories. Their evidence does not establish direct purified binding, an exclusively nuclear location or an isoform-specific OpenCell interaction.
The membrane organization, protein-partitioning and APP/plasminogen studies were inspected in the original, including mouse and human constructs. Existing process assertions are refined where necessary to their regulatory roles. Hormone/developmental transfers and the membrane-proteomics target assignment remain undecided where original evidence could not be read. A plausible cholesterol phenotype is not used to invent detailed assay results or dismiss a curator's assertion.
The source's duplicate catalytic cores and repeated reference findings were replaced by the supported synthesis and reference-only evidence links. Two short exact core quotations are used once each: nine words from PMID11519011 and thirteen from PMID15577914. This avoids repeated long excerpts; citation count is not quotation count. The missing PMID22010141 reference entry is restored, and later topology PMID27919032 is cited without rewriting either source cache. The two official 516/475-residue products are documented without extending full-length construct results to the shorter isoform.
Prior provenance is retained: PR3245 approved the obsolete-term migration, which remains represented in the same three source rows. PR7 supplied a bot narrative, PR119 had a COMMENTED review, and PR3214 supplied broad import approval. These are not represented as a DHCR24-specific human sign-off. Existing Falcon research, pathway, source files and the prior history record remain immutable. The historical COMPLETE scalar was not completion credit; ordinary status computation determines the final status for this augmentation.
This appendix supersedes the earlier undecided membrane-proteomics verdict. Membrane association is established independently of the unavailable YTS peptide assignment, so the broad GO:0016020 row now proposes GO:0005789. The HDA source identity is preserved, and its unresolved fraction is not presented as an ER-specific measurement. Under the explicit instruction to use a supported specific replacement for an overly broad term, the action is MODIFY. Counts are now 14 ACCEPT, 8 KEEP_AS_NON_CORE, 22 MODIFY and 6 UNDECIDED; all 50 source identities, two products and two core activities remain.
The p53-binding unit is retained because human recombinant binding/competition and separation-of-function experiments distinguish it from sterol reduction. It does not assign broad stress phenotypes as core processes. A knowledge gap now states that the endogenous interaction compartment and route from the ER-associated pool are unresolved; the reported nuclear observation is not substituted for an interaction-localization experiment.
Short exact anchors now support the topology, membrane localization, DHCR7 association and APP/plasminogen claims. The evidence guide distinguishes recovered original Results from two caches that omit those passages despite their full-text flags. Quotations are used once with reference-only support elsewhere, within 25 words per paper across the review and guide. The previous notes remain unchanged. The biological description no longer contains product-evidence provenance; those boundaries remain in the product descriptions and guide.
The two skin-development transfers remain undecided: the accessible abstract establishes the epidermal phenotype, but the original developmental experiments and controls remain unavailable. Recovery of those experiments would resolve the present evidential boundary. The plasminogen and APP replacements retain their original source fields and describe measured regulatory effects of membrane organization; they do not assign protease activities. The DHCR7 cellular effect remains supported, while its division among protein abundance, sterol flux and allostery is unresolved.