ACVR1B manual research and annotation audit, 2026-10-25

Human ACVR1B (P36896), also ALK4/ActRIB/SKR2, was seeded before review. All 81 existing rows have been reviewed, preserving original term, evidence, source, qualifiers and supporting-entity fields. No NEW annotation was added.

Research provenance

The parent attempted Perplexity and Falcon research. /tmp/aigr-activin-ACVR1B-deep.log records Perplexity HTTP 401 insufficient_quota; /tmp/aigr-activin-ACVR1B-falcon.log records the Falcon authentication failure and failed Perplexity fallback. Providers were not retried. This file is the manual research record; no provider-named report was fabricated.

All original PMID caches were read, along with the curated UniProt function, domain, modification and binding-site statements and available Reactome entries. Full text of PMID:21356369 was inspected specifically because its title emphasizes ALK7 while the annotation concerns ALK4. Target-specific mammalian nervous-system evidence was added from PMID:31676717, including the fetched full text. Targeted attempts to access full publisher text of PMID:18039968 and PMID:7890768 returned unavailable/403 responses; specific unresolved claims remain UNDECIDED.

Mechanistic synthesis and boundaries

ALK4 is a type-I membrane serine/threonine receptor kinase, not a soluble transcription factor. Type-II activin receptors first bind activin and recruit ALK4; phosphorylation of the ALK4 GS domain activates its downstream kinase signaling PMID:8622651. The paper states: “ActR-IB acts as a downstream transducer of activin signals.” It also explicitly says that isolated ActRIB does not bind ligand on its own.

This is compatible with direct ligand contact within the assembled complex. Extracellular ALK4 mutations alter radioligand cross-linking in the presence of ActRII PMID:12665502. The contributes_to receptor and activin-binding rows retain their weaker complex-component interpretation; no independent high-affinity ligand capture was invented.

Activated ALK4 binds R-SMADs SMAD2/3, which subsequently leave the receptor and assemble with SMAD4 PMID:9892009. SMAD7 and FKBP12 regulate receptor signaling, with recruitment of SMURF1 controlling ubiquitination PMID:16720724. SMAD4 binding in a protein interaction screen is not evidence that it is a receptor substrate. Receptor ubiquitination does not prove ubiquitin-binding activity.

Nodal also uses ALK4, with an EGF-CFC coreceptor such as CRIPTO [PMID:11389842; PMID:12052855]. Activin does not share this coreceptor requirement PMID:14966532. Myostatin can use ActRIIB with ALK4 or ALK5 PMID:14517293. BMP-3b/GDF10 is functionally facilitated by ALK4/ActRIIA and activates SMAD2/3, despite the BMP name PMID:22155034. Neither the kinase-family name nor shared type-II receptors establish identical ligand repertoires. SB-431542 inhibits ALK4, ALK5 and ALK7 and therefore does not uniquely implicate ALK4 by itself PMID:12065756.

Biological outputs

• Human K562 experiments show activin-dependent hemoglobin accumulation and reduced proliferation PMID:9032295. Truncated dominant-negative ALK4 isoforms block the growth arrest and G1 accumulation PMID:11117535. The G1/S annotation is refined to negative regulation, not progression through the transition.
• In the full text of PMID:21356369, ALK4 knockdown reduces trophoblast invasion whereas ALK7 or Nodal knockdown increases it. Thus the ALK7-focused title does not invalidate the positive ALK4 migration annotation; the authors connect the ALK4 effect to activin signaling.
• Mouse conditional Alk4 deletion supports development of cortical somatostatin interneurons PMID:31676717. This corroborates a gene-specific mammalian role without asserting a direct human neural experiment or identical functions for every member of the broad PAINT family.
• Myostatin/BMP-3b studies support negative adipocyte/osteoblast differentiation contexts, while ALK4 promotes erythroid differentiation elsewhere. Both signs are cell-context dependent, not mutually exclusive universal claims. Mouse HS-72 hybridomas support an activin-triggered apoptotic output PMID:9884026.

Unresolved evidence

The direct inhibin-binding row (GO:0034711) remains UNDECIDED. PMID:7890768 explicitly assigns inhibin binding to ActRII and reports failure of the ActRII-SKR2/ALK4 complex to form. InhBP association in PMID:11266516 does not prove direct ALK4-inhibin binding. Full text was inaccessible, so this experimental annotation was not removed on an incomplete source.

The peptidyl-threonine phosphorylation row (GO:0018107) attached to PMID:18039968 also remains UNDECIDED: the abstract describes TTRAP/Alk4 signaling and zebrafish development, but the receptor construct and precise phosphorylation assay could not be verified directly. The receptor's general kinase and autophosphorylation functions have independent biochemical/curated support.

The legacy Reactome R-NUL-1181349 reference is unresolved; it was not silently reassigned to a guessed human identifier. Its plasma-membrane assertion is independently supported by primary receptor-binding experiments.

Review coverage

Final decisions: MODIFY 31, ACCEPT 28, REMOVE 12, KEEP_AS_NON_CORE 8, UNDECIDED 2. All 81 rows have decisions; no PENDING or NEW rows. Bare protein binding is never ACCEPT; seven such rows are refined to demonstrated receptor/SMAD binding classes and twelve are removed as uninformative, not biologically false.

Two core functions distinguish the specific activin type-I receptor activity from receptor kinase transduction in the nodal/CRIPTO context. Downstream cellular phenotypes remain non-core.

Annotation audit ledger

Rows are one-based YAML order.

Validation: just validate human ACVR1B passed without warnings after the initial review. just backfill-record-sections genes/human/ACVR1B derived UniProt record sections. Final follow-up validation and history validation are recorded with the audit history.

Final follow-up: just validate human ACVR1B passed with zero warnings, git diff --check passed for ACVR1B/SMAD3, and the scaffolded ACVR1B history record validated without issues.