Paired reviews 01–10

The initial pass assesses all 688 existing GOA rows across these ten human–horse pairs, with no PENDING actions. All ten human genes have completed Falcon research reports; GHSR also has a horse-specific report because direct horse literature warranted it. Primary sources are cached separately, and each horse target has a reproducible alignment with mapped human sequence features. Gene reviews remain IN_PROGRESS where assay, phylogenetic or protein-model questions are unresolved.

Gene	Human / horse GOA rows	External prediction assessment	Main finding
VAPA	182 / 3	PLI, function paragraph	ER membrane adaptor with an MSP domain does not acquire nematode sperm-crawling function.
CAPSL	3 / 1	NPI, function paragraph	Venom-secretion narrative conflicts with intracellular family biology; human endothelial studies and emerging mammalian ciliary evidence inform function.
MYL10	12 / 5	NPI, function paragraph	Venom narrative is unsupported for the conserved EF-hand light-chain protein.
KRIT1	51 / 5	NPI, function paragraph	Rap1-associated CCM junctional scaffold is assigned a piRNA/germ-granule narrative.
CTDSP2	20 / 0	UNC, GO:0016301 kinase activity	The selected 174-residue horse model lacks the canonical phosphatase catalytic region. A gene-name-based phosphatase/kinase contradiction would overstate exact-sequence evidence.
PEA15	21 / 4	UNC, GO:0008643 carbohydrate transport	Direct human experiments establish regulation of glucose-transporter recruitment and uptake; whether the exact transport-process term is appropriate remains unresolved.
ALDH7A1	42 / 10	UNC, GO:0043878 substrate-specific dehydrogenase	Broad aldehyde activity does not establish glyceraldehyde-3-phosphate oxidation; an internal 18-residue horse deletion adds uncertainty.
GHSR	90 / 33	PLI, function paragraph	Ghrelin receptor is assigned oxytocin-receptor function. A separate GHRH-receptor GO error propagates from human to horse and is corrected in both reviews.
ALG5	19 / 2	PLI, function paragraph	ALG5-family UDP-glucose chemistry is confused with GDP-mannose/DPM synthesis; the selected horse protein has an internal 30-residue deletion.
HSPD1	178 / 7	PLI, function paragraph	Mitochondrial type I Hsp60 is assigned type II CCT/TRiC and its clients. A concordant horse electronic CCT-complex annotation is removed. A 32-residue deletion limits exact Hsp60 activity transfer.

The three GO predictions have standard prediction-review YAMLs. The seven original function paragraphs are preserved verbatim in gene-specific GENE-protnlm-function-review.md reports and assessed as atomic claims. Paragraph assessments describe the principal biological assertion; shared generic clauses are separately identified and do not rescue an incorrect specific complex, substrate or physiological context. Training-set membership is unknown.

Human GHSR and horse GHSR receive particular attention to the distinction between ghrelin and growth-hormone-releasing hormone. PMID:10604470 explicitly separates the pathways. Horse-specific PMID:42426797 supports sequence-confirmed receptor transcript in pituitary and adrenal medulla; PMID:29526750 studies receptor-associated ghrelin responses in equine chondrocytes. Tissue expression, pharmacology and exact-accession binding competence remain distinct evidence claims.

Validation: all 20 main gene YAMLs pass individual validation, all three prediction YAMLs pass schema and evidence validation, all seven original paragraphs match the frozen source CSV, and all 20 HTML pages render. Matching history records were scaffolded and validated for each curated gene. Warnings about omitted core functions for unresolved horse models and uninspected details of particular IBA placements remain explicit; no unsupported function was added merely to silence an advisory.