ProtNLM2 External predictions
View prediction review YAML · MTMR9-protnlm-predictions-review.yaml · Review status: COMPLETE
MTMR9 regulatory, complex, and negative-autophagy predictions are supported; specific stabilization and protein-phosphatase-binding transfers remain uncertain for the shortened horse sequence.
Source documents: projects/PROTNLM_EVALUATION/mammal-benchmark/horse40-predictions.csv · projects/PROTNLM_EVALUATION/mammal-benchmark/predictions.jsonl.gz · projects/PROTNLM_EVALUATION/mammal-benchmark/paired-sequences/MTMR9.json · genes/HORSE/MTMR9/MTMR9-hypotheses/horse40-negative-autophagy-regulation/openscientist.md
Review score: 2 = concordant with evidence; 1 = uncertain; 0 = discordant with evidence. This is an assessment score, not a model probability.
COR — Correct novel prediction Review score: 2/2
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-09-08
Review rationale: Human MTMR9 directly complexes with and regulates MTMR6/MTMR8; catalytic inactivity of MTMR9 is compatible with this regulatory or complex-associated claim. The horse protein is 97.8% identical over 507 aligned residues, retaining the C-terminal association region. The 42-residue deletion at the GRAM/phosphatase-domain junction is a gene-model caveat, but does not remove the entire partner-binding region. This is an inferred conserved regulatory role, new relative to the frozen horse GOA, rather than evidence for intrinsic phosphatase catalysis.
Supporting Evidence:
- PMID:22647598: "Complex formation between the active myotubularins and MTMR9 increases their catalytic activity and alters their substrate specificity"
- file:HORSE/MTMR9/MTMR9-bioinformatics/RESULTS.md: "The selected horse sequence lacks human residues 98–139, crossing the GRAM-domain/myotubularin-domain junction; the C-terminal region implicated in partner association is retained. Complex formation and downstream regulation remain inferences with a gene-model caveat."
COR — Correct novel prediction Review score: 2/2
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-09-08
Review rationale: Human MTMR9 can contribute to negative regulation of autophagy as a catalytically inactive activating subunit of the MTMR8-MTMR9 phosphoinositide-phosphatase complex. The OpenScientist audit confirmed that the horse sequence is a 1:1 MTMR9 ortholog with an identical inactive P-loop and a conserved C-terminal coiled-coil dimerization region, resolving the 42-residue deletion as a gene-model caveat rather than loss of the partner-dependent mechanism. This role should be framed as a contributes_to, MTMR8-MTMR9-complex-mediated, context-dependent, non-primary autophagy effect; MTMR9 knockdown does not uniformly enhance autophagic clearance in neurons. The GO:0010507 prediction is supported by mammalian transfer, not direct horse evidence or an intrinsic lipid-phosphatase activity.
Supporting Evidence:
- PMID:22647598: "the MTMR8/R9 complex inhibits autophagy."
- file:HORSE/MTMR9/MTMR9-bioinformatics/RESULTS.md: "The selected horse sequence lacks human residues 98–139, crossing the GRAM-domain/myotubularin-domain junction; the C-terminal region implicated in partner association is retained. Complex formation and downstream regulation remain inferences with a gene-model caveat."
- file:HORSE/MTMR9/MTMR9-hypotheses/horse40-negative-autophagy-regulation/openscientist.md: "The **PTP-like pseudophosphatase domain** (P-loop `LIHGTEGTDSTLQVT`) and the **C-terminal coiled-coil dimerization region** are fully conserved"
UNC — Uncertain Review score: 1/2
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-09-08
Review rationale: Human MTMR9 association stabilizes its partners, but protein stabilization is a specific downstream consequence of productive association. It is not established merely by sequence similarity, annotation overlap, or the domain conservation that supports complex-mediated autophagy transfer. Direct evidence that this shortened horse model retains partner stabilization is needed.
Supporting Evidence:
- PMID:19038970: "the formation of this heteromer in cells resulted in increased protein levels of both MTMR6 and MTMR9, probably due to the inhibition of degradation of both proteins."
- PMID:22647598: "Complex formation between the active myotubularins and MTMR9 increases their catalytic activity and alters their substrate specificity"
- file:HORSE/MTMR9/MTMR9-bioinformatics/RESULTS.md: "The selected horse sequence lacks human residues 98–139, crossing the GRAM-domain/myotubularin-domain junction; the C-terminal region implicated in partner association is retained. Complex formation and downstream regulation remain inferences with a gene-model caveat."
COR — Correct novel prediction Review score: 2/2
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-09-08
Review rationale: Human MTMR9 directly complexes with and regulates MTMR6/MTMR8; catalytic inactivity of MTMR9 is compatible with this regulatory or complex-associated claim. The horse protein is 97.8% identical over 507 aligned residues, retaining the C-terminal association region. The 42-residue deletion at the GRAM/phosphatase-domain junction is a gene-model caveat, but does not remove the entire partner-binding region. This is an inferred conserved regulatory role, new relative to the frozen horse GOA, rather than evidence for intrinsic phosphatase catalysis.
Supporting Evidence:
- PMID:22647598: "Complex formation between the active myotubularins and MTMR9 increases their catalytic activity and alters their substrate specificity"
- file:HORSE/MTMR9/MTMR9-bioinformatics/RESULTS.md: "The selected horse sequence lacks human residues 98–139, crossing the GRAM-domain/myotubularin-domain junction; the C-terminal region implicated in partner association is retained. Complex formation and downstream regulation remain inferences with a gene-model caveat."
UNC — Uncertain Review score: 1/2
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-09-08
Review rationale: Human MTMR9 directly binds the lipid phosphatases MTMR6/7/8, but this prediction names protein phosphatase binding. The interaction evidence should not be conflated with a phosphoprotein-substrate partner class. A verified protein-phosphatase interaction or clarified ontology scope is needed before accepting this exact MF; the shortened horse domain junction adds a transfer caveat.
Supporting Evidence:
- PMID:22647598: "Complex formation between the active myotubularins and MTMR9 increases their catalytic activity and alters their substrate specificity"
- file:HORSE/MTMR9/MTMR9-bioinformatics/RESULTS.md: "The selected horse sequence lacks human residues 98–139, crossing the GRAM-domain/myotubularin-domain junction; the C-terminal region implicated in partner association is retained. Complex formation and downstream regulation remain inferences with a gene-model caveat."
CNN — Correct but not novel Review score: 2/2
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-09-08
Review rationale: Enzyme regulation is supported by human MTMR9 stimulation of MTMR6 and MTMR8 in biochemical experiments and the highly conserved horse sequence with retained C-terminal association region. The exact broad term already occurs in the frozen horse GOA. The internal 42-residue deletion leaves a caveat about the magnitude and substrate specificity of regulation in this selected gene model; it does not turn this regulator prediction into a catalytic claim.
Supporting Evidence:
- PMID:22647598: "Complex formation between the active myotubularins and MTMR9 increases their catalytic activity and alters their substrate specificity"
- file:HORSE/MTMR9/MTMR9-bioinformatics/RESULTS.md: "share 97.8% identity among 507 paired residues. Paired coverage is 92.3% of human and 100.0% of horse."
COR — Correct novel prediction Review score: 2/2
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-09-08
Review rationale: Human MTMR9 directly complexes with and regulates MTMR6/MTMR8; catalytic inactivity of MTMR9 is compatible with this regulatory or complex-associated claim. The horse protein is 97.8% identical over 507 aligned residues, retaining the C-terminal association region. The 42-residue deletion at the GRAM/phosphatase-domain junction is a gene-model caveat, but does not remove the entire partner-binding region. This is an inferred conserved regulatory role, new relative to the frozen horse GOA, rather than evidence for intrinsic phosphatase catalysis.
Supporting Evidence:
- PMID:22647598: "Complex formation between the active myotubularins and MTMR9 increases their catalytic activity and alters their substrate specificity"
- file:HORSE/MTMR9/MTMR9-bioinformatics/RESULTS.md: "The selected horse sequence lacks human residues 98–139, crossing the GRAM-domain/myotubularin-domain junction; the C-terminal region implicated in partner association is retained. Complex formation and downstream regulation remain inferences with a gene-model caveat."