GAD1

Summary: The verified GAD family ancestral placement supports glutamate decarboxylase activity.

Reason: GAD67 conserves the PLP-dependent conversion of L-glutamate to GABA and carbon dioxide, directly supported by human full-length activity in PMID:10671565 and the structural/functional evidence in PMID:17384644. The actual PAINT catalytic assertion lies on the verified human target path, with incoming, literal and main donors agreeing and no target-specific loss demonstrated. A target self-donor is legitimate experimental grounding. This conserved full-length activity does not imply activity of the experimentally inactive GAD25 product.

Summary: The original experiment measures active full-length GAD67 alongside inactive GAD25.

Reason: The selected original Methods, Results and Figure 4 caption explicitly compare equimolar translated full-length GAD67 and GAD25 using radiolabeled glutamate-to-CO2 activity, PLP supplementation, methionine normalization and translation controls. Full-length GAD67 is active, whereas GAD25 is not. Thus the paper title and variant result do not invalidate this gene-level canonical glutamate decarboxylase annotation. The inactive short product remains separately documented.

Summary: Human GAD1 has directly supported glutamate decarboxylase activity.

Reason: The authentic abstract and publisher preview describe the catalytic distinction between constitutively active GAD67 and more readily inactivated GAD65. Official PDB 2OKJ identifies human GAD1/Q99259, distinct from the GAD2 structure 2OKK. The GAD1 crystallization entity reconstructs as residues 93–594 plus two terminal histidines; it is an engineered construct rather than a natural splice product. Complete subscription Methods were not independently obtained, but the accessible target-specific catalytic evidence and independent full-length assay support this annotation.

Summary: The enzyme and ortholog mappings agree with direct human GAD67 catalysis.

Reason: The incoming EC/Rhea and mouse ortholog support describes the established conversion of L-glutamate to GABA and carbon dioxide. Human full-length GAD67 is directly active in the controlled translation/CO2 assay of PMID:10671565; PMID:17384644 independently establishes the catalytic GAD1 enzyme. The actual mouse Gad1 knockout study PMID:9177246 measures greatly reduced brain GAD activity and GABA, corroborating the donor chemistry. These converging data support the mapping without treating every natural splice product as catalytically equivalent.

Summary: The cited GAD25 study does not resolve the exact GAD1–GAD2 interaction assay.

Reason: The original GAD67/GAD25 activity sections were read, but their discussion of GAD65 association and microvesicle targeting is tentative and cites earlier work. No direct GAD1–Q05329 binding experiment was identified in the inspected sections. The official Reactome model independently includes a GAD1:GAD2 heterodimer, which makes the biological association plausible but does not recover this source-specific IPI experiment. Retain uncertainty about the cited assay rather than declaring the interaction false or converting a tentative statement into a direct heterodimer-binding result.

Summary: A human interaction screen records GAD1 binding to HAAO.

Reason: The authentic abstract and exact current IntAct target record join Q99259 to HAAO/P46952 in pooled yeast two-hybrid testing. The paper-level coaffinity validation rate cannot be assigned to this individual pair. The positive pooled two-hybrid interaction is preserved as evidence. Remove the generic protein-binding annotation because it supplies no informative GAD1 function; this does not deny the interaction or establish a catalytic HAAO partnership or endogenous complex.

Summary: The GAD1 peptide binds NCK1 SH3 domains in the actual domain screen.

Reason: The retained publisher Table S2 contains two NCK1_HUMAN–DCE1_HUMAN entries. GAD1 supplies the exact canonical peptide QSLRGVPDSPQRR at residues 520–532, while NCK1/P16333 supplies the SH3 domains; current interaction records identify NCK1 regions 2–61 and 115–165. Therefore SH3 domain binding is the supported informative refinement of generic protein binding. The domain orientation must not be reversed, and peptide/domain assays do not establish binding of every splice product or an endogenous full-length complex.

Proposed replacements: SH3 domain binding

Summary: Remove uninformative bare protein binding (unresolved GAD1-AGTRAP join).

Reason: Bare protein binding (GO:0005515) carries no functional information and is removed per the protein-binding policy; the cited full text is readable, so the term is uninformative rather than undecidable, and removal does not resolve or deny the unresolved GAD1-AGTRAP partner join.

Summary: The original binary interaction table supports the GAD1–HAAO pair.

Reason: Genuine Supplementary Table 2G includes the exact GAD1–HAAO/P46952 positive pair, corroborated by the current curated interaction records. The recovered normal text contains Introduction and Discussion rather than complete Methods, so the target table and recorded two-hybrid assay scope are stated explicitly. The positive construct-level interaction remains supported, but remove the uninformative generic protein-binding annotation. Neither a catalytic HAAO partnership nor an endogenous complex is established, so no specific replacement is invented.

Summary: The original binary interaction table supports GAD1 binding to STK3.

Reason: Genuine Supplementary Table 2G includes the exact GAD1–STK3/Q13188 positive pair. Current interaction records report the associated two-hybrid screening and validation forms. These records do not independently show that STK3 phosphorylates GAD1 or that GAD1 regulates a Hippo pathway step. Preserve the tested GAD1–STK3 interaction, source and partner as positive evidence, but remove the uninformative generic protein-binding annotation; no specific kinase-regulatory function is established.

Summary: The original binary interaction table supports GAD1 binding to CMTM5.

Reason: Genuine Supplementary Table 2G includes the exact GAD1–CMTM5/Q6RW13 positive pair. The curated two-hybrid records corroborate the pair, but no membrane recruitment, vesicle function or endogenous interaction mechanism is established by this screen. Preserve the positive GAD1–CMTM5 pair without importing a splice suffix from another study. Remove the generic protein-binding annotation as uninformative, rather than denying the interaction or inventing a membrane-related molecular function.

Summary: The original binary interaction table supports GAD1 binding to AGTRAP.

Reason: Genuine Supplementary Table 2G includes the exact GAD1–AGTRAP/Q96DZ9 positive pair. This independently supports the interaction, while leaving the different PMID:21516116 attribution unresolved. The tested two-hybrid interaction does not establish GAD1 participation in angiotensin-receptor regulation, and no partner isoform suffix is added to this original context. Remove the uninformative generic protein-binding annotation while preserving this positive AGTRAP interaction and the separate source-attribution uncertainty.

Summary: The isoform interaction study supports the reference GAD1–CMTM5-2 pair.

Reason: The genuine clone table translates the study reference GAD1_1 clone to the exact current 594-residue Q99259-1 sequence, and Table S2B records its positive CMTM5 interaction. The incoming Q6RW13-2 participant is preserved literally. The Gal4 isoform screen used retesting and sequence confirmation. Study GAD1_2 and GAD1_3 are separate experimental clone names, not automatically current UniProt products; selected alternative-clone PCA results are not assigned to this reference pair. Preserve this tested CMTM5 interaction and its literal participant suffix, but remove the generic protein-binding annotation because it does not identify an informative molecular function; the positive assay is not rejected.

Summary: The isoform interaction study supports reference GAD1 binding to MAL2.

Reason: The exact 594-residue reference GAD1_1 clone is positive for MAL2/Q969L2 in genuine Table S2B. Other tested GAD1 clones differ in interaction patterns, so the source cannot establish uniform binding across all current products. The screen/retests support this construct-specific binary interaction without proving a physiological vesicle-targeting or transcytosis role for GAD1. Remove the uninformative generic protein-binding annotation while preserving the positive MAL2 pair and product limitations.

Summary: The isoform interaction study supports the reference GAD1–AGTRAP-2 pair.

Reason: The genuine Table S2B records reference GAD1_1 binding to AGTRAP; its translated sequence is exactly canonical Q99259-1. Preserve incoming AGTRAP Q96DZ9-2 and the assay-specific clone context. Study alternative clone numbers must not be substituted for the undescribed current Q99259-2 product. Preserve the positive AGTRAP interaction without assigning receptor-regulatory function or selected alternative-clone PCA evidence to this pair. Remove the generic protein-binding annotation as uninformative, not because the tested interaction is false.

Summary: GAD1 remains an interaction partner for the tested HAAO variants.

Reason: The original Y2H Methods and genuine Supplementary Data 2 rows 476–477 show HAAO/P46952 T42S and I37V tested against GAD1/2571, with Disruption=0. These variant positions belong to HAAO, not GAD1. The controlled human wild-type/mutant two-hybrid comparison supports retained binding in the tested constructs; no pair-specific PCA result or endogenous metabolic complex is inferred. Preserve the positive construct-level HAAO interaction and its variant ownership. Remove the uninformative generic protein-binding annotation without denying binding or inventing a metabolic molecular function.

Summary: HuRI records the GAD1-1–HAAO-1 interaction.

Reason: The authentic HuRI table contains Q99259-1 paired with P46952-1 in array, prey-pooling and validated two-hybrid records. Those related screening forms are not three independent physiological replications. Preserve the incoming P46952 partner and tested P46952-1 clone as positive interaction evidence. Remove the generic protein-binding annotation as uninformative; no catalytic HAAO partnership is inferred and the assay result is not denied.

Summary: HuRI records GAD1-1 binding to CMTM5-2.

Reason: The authentic HuRI extract joins Q99259-1 to the literal Q6RW13-2 participant across the recorded array, prey-pooling and validated two-hybrid forms. This preserves the partner suffix and tested GAD1 clone. The screen supports binary binding but does not independently establish an endogenous membrane or vesicle function for the enzyme. Remove the uninformative generic protein-binding annotation while retaining the positive CMTM5 pair and exact clone/suffix context.

Summary: HuRI records GAD1-1 binding to MAL2.

Reason: The authentic HuRI extract joins Q99259-1 to MAL2/Q969L2 with the recorded two-hybrid screening and validation forms. The positive MAL2 pair remains supported, without treating the three record forms as separate in-vivo replications or assigning GAD1 a transcytosis function. Remove the generic protein-binding annotation because it supplies no informative molecular function; this is not a rejection of the interaction. Current alternative GAD1 products were not thereby shown to share this interaction.

Summary: HuRI records GAD1-1 binding to AGTRAP-2.

Reason: The authentic HuRI extract joins Q99259-1 to literal Q96DZ9-2. Array, prey-pooling and validated two-hybrid records support the tested binary interaction, not an endogenous receptor-regulation mechanism. Preserve the partner suffix, reference GAD1 clone and positive AGTRAP interaction. Remove the uninformative generic protein-binding annotation without denying the assay result or inventing receptor-regulatory activity.

Summary: GAD67 acts in the cytoplasm, with additional specialized neuronal pools.

Reason: Actual GAD67 antibody staining in the selected original PMID:15347806 places protein in neuronal cytoplasm, and the verified broad PAINT cytoplasm assertion at PTN000242648 lies on the exact human sequence path. The current incoming donor set contains an additional worm entry relative to the retained literal/main assertion; both dated contexts remain distinct. Neither that difference nor the broad ancestral placement establishes a defect. Cytoplasm is nonexclusive and does not exclude protein at neuronal terminals or associated vesicle membranes.

Summary: The cytoplasmic mouse-to-human transfer is supported by actual GAD67 immunostaining.

Reason: The complete donor capture identifies PMID:15347806 and 20096683. In the selected original 15347806 Results/Figure 2, anti-GAD67 produces cytoplasmic staining distinct from a nuclear Npas reporter, providing target-protein evidence. The second donor paper was not read beyond its abstract, and is not represented as another independently inspected assay. Broad cytoplasm is well supported without asserting exclusive soluble cytosol localization.

Summary: The GAD1 vesicle participant does not establish the event-level plasma-membrane assignment.

Reason: The official R-HSA-888589 hierarchy includes GAD1 through a loaded-vesicle GAD1:GAD2 assembly. Its Q99259 component R-HSA-888580 is explicitly assigned to the GABA transport-vesicle membrane; the release summary assigns plasma-membrane locations to SNAP25 and syntaxin. This participant-resolved model supports the vesicle pool but does not establish plasma-membrane localization of GAD1 itself. Mark this source-specific event-to-participant compartment transfer as over-annotated. This does not deny genuine event membership or establish that GAD1 can never associate with the cell surface.

Summary: The specific cell-cortex donor assay remains unresolved.

Reason: The actual mouse P48318 IDA record cites PMID:11064363 and carries neuronal/anatomical extensions. The accessible septin-focused abstract does not expose the GAD67 cell-cortex staining or its spatial resolution. That is an access gap, not proof of a wrong gene. Broad cytoplasmic and presynaptic GAD67 evidence does not by itself establish the specialized subplasmalemmal cortex compartment defined by this term; preserve the original donor context and uncertainty.

Summary: Full-length GAD67 directly consumes L-glutamate in GABA synthesis.

Reason: The selected original full-length GAD67 versus GAD25 experiment measures radiolabeled glutamate-to-CO2 conversion, demonstrating the chemical breakdown step rather than merely a change in neurotransmitter abundance. Full-length GAD67 is active and GAD25 inactive. This directly supports L-glutamate catabolism by the characterized enzyme without implying all routes of glutamate disposal.

Summary: The catalytic GAD67 study supports the entry step, not every reaction of the GABA shunt.

Reason: The accessible target-specific structural/functional evidence establishes sustained GABA production by human GAD67. Glutamate decarboxylation is explicitly the first reaction in the GO-defined shunt, whereas transamination and succinate-semialdehyde oxidation are performed by other enzymes. Retain this broader pathway association as non-core, without converting substrate production into evidence for all downstream catalytic activities or measured in-vivo shunt flux.

Summary: GAD1 catalyses the entry reaction of the broader GABA shunt.

Reason: The official GO definition explicitly begins the shunt with glutamate decarboxylase converting glutamate to GABA, followed by other enzymes producing succinate. The actual mouse donor PMID:9177246 directly measures Gad1-dependent GAD activity and GABA, and human enzyme assays independently establish this first chemical step. Retain the broader pathway participation as non-core; these data do not quantify human GAD1-dependent flux through the complete succinate-producing route.

Summary: Cloned human GAD67 supplies the first chemical step of the GABA shunt.

Reason: The authentic PMID:1549570 abstract explicitly identifies separate human GAD65 and GAD67 cDNAs and enzymatically active bacterial expression products. This supports the glutamate-to-GABA step named in the official shunt definition, with independent human activity corroboration. Retain participation in this broader pathway as non-core rather than claiming that this gene-identity study measured the complete GABA-to-succinate pathway or its physiological flux.

Summary: The complete cited evidence for the specific transmission process remains unavailable.

Reason: The authentic abstract reports mouse Gad2/GAD65 knockout, preserved GAD67 expression and total GABA, and increased seizure susceptibility. It therefore includes GAD67 context and cannot be rejected merely for foregrounding GAD65. GAD1 certainly supplies the GABA transmitter, but the accessible abstract does not resolve its own participation in the release, receptor-activation or electrical-response steps named by the GO term. The complete source-specific experiments were not recovered. Retain UNDECIDED for that access gap, without concluding that GAD1 cannot contribute to transmission or treating transmitter supply alone as the missing experiment.

Summary: Conserved glutamate decarboxylation directly produces GABA.

Reason: GAD67 performs the GABA-forming chemical step, established by human enzyme assays and the direct mouse Gad1 activity/GABA experiment. The verified catalytic IBD at PTN002903772 lies on the exact Q99259 target path. Incoming donors are retained separately from the later literal/main list, which additionally names mouse MGI:95632; that dated difference is not a biological defect. No target-specific loss of GABA synthesis is demonstrated.

Summary: The actual mouse donor and human assays support direct GABA biosynthesis.

Reason: The P48318 donor is tied to PMID:9177246, whose original Methods/Results measure GAD activity and GABA after Gad1 disruption. Human full-length GAD67 independently catalyses glutamate decarboxylation. This is direct chemical participation in GABA production, not a process assignment based only on a nonspecific developmental phenotype, and it need not imply activity of every splice product.

Summary: Vesicle-associated GAD67 is a supported non-core pool, consistent with the four specific transport-vesicle annotations.

Reason: The selected original PMID:10671565 Discussion suggests GAD65 association and microvesicle targeting, but does not supply a newly inspected binding or localization assay. Independent official Reactome component R-HSA-888580 explicitly places human Q99259 at the GABA transport-vesicle membrane, both in a GAD1 homodimer and in GAD1:GAD2 assemblies. Retain this broad location as non-core for the same modeled pool underlying the four GO:0061202 annotations. This does not establish exclusively heterodimer-mediated recruitment, exclude broader cytoplasmic or presynaptic GAD67, or assign equivalent targeting to inactive GAD25.

Summary: The broad carbon-carbon lyase mapping can be resolved to glutamate decarboxylase activity.

Reason: The IPR002129 mapping identifies the appropriate enzyme class, but direct human full-length assays and GAD1 structural/functional evidence identify the specific L-glutamate-to-GABA reaction. GO:0004351 is a verified is_a descendant of GO:0016830. Refine the broad class without assigning this activity to every splice product.

Proposed replacements: glutamate decarboxylase activity

Summary: The carboxy-lyase mapping can be refined to the established glutamate reaction.

Reason: The IPR021115-derived carboxy-lyase class agrees with removal of the glutamate carboxyl group as CO2. Human GAD67 activity establishes the specific substrate and product, so GO:0004351 is the informative verified is_a descendant of GO:0016831. The refinement concerns the characterized enzyme, not the inactive GAD25 variant.

Proposed replacements: glutamate decarboxylase activity

Summary: Pyridoxal phosphate is the cofactor of the GAD67 catalytic activity.

Reason: The domain mapping is corroborated by the original human enzyme literature: PMID:17384644 identifies GAD67 as a PLP-dependent enzyme and distinguishes its stable catalytic configuration from GAD65 cofactor loss; the controlled GAD67/GAD25 assay in PMID:10671565 includes PLP. Retain cofactor binding as part of the same decarboxylase mechanism, without constructing a separate generic binding core or extending cofactor competence to all products.

Summary: GAD67 is deployed in inhibitory neuronal axons and terminals.

Reason: The complete mouse donor capture ties axon IDA to PMID:12074840 and 15479642. The 12074840 original preview confirms a double-staining marker design, although its GAD67-specific panel remains unavailable; 15479642 remains abstract-only. Independent original GAD67 protein labelling of inhibitory presynaptic terminals corroborates the conserved axonal deployment. Retain the location without inventing a donor antibody protocol or importing neuronal/anatomical extensions.

Summary: The human GAD1 structure supports a homodimer rather than unspecified identical-protein binding.

Reason: The curated UniProt SUBUNIT statement, backed by experimental evidence from PMID:17384644, identifies a homodimer. Independently, official PDB 2OKJ identifies human GAD1/Q99259 and an author/PISA-supported homodimer. Its engineered 504-residue entity equals Q99259 residues 93–594 plus two terminal histidines; it is distinct from the GAD2 entry 2OKK and from natural splice products. Thus GO:0042803 protein homodimerization activity is the verified specific descendant of identical protein binding. Do not infer the same assembly for the short inactive GAD25 product.

Proposed replacements: protein homodimerization activity

Summary: GAD67 protein occurs in inhibitory axon terminals.

Reason: The original source of the mouse axon-terminus IDA is PMID:11064363, for which the exact GAD67 panel remains uninspected. Independent GAD67 antibody Methods and presynaptic-terminal quantification in PMID:23258346 directly establish protein in inhibitory terminal regions, supporting the whole localization claim. Retain this broad terminal placement without pretending that the septin paper’s specific assay was recovered or transferring its anatomical extensions.

Summary: The neuron-projection-terminus transfer agrees with direct terminal GAD67 evidence.

Reason: The actual P48318 IDA donor points to PMID:12074840, whose relevant marker-specific panel was not accessible. Its double-staining design and curated assertion are retained, with independent GAD67-positive inhibitory terminal measurements in PMID:23258346 corroborating the broader location. This is a localization claim rather than a new requirement for GAD1 in projection development, and the mouse neuron extension is not imported.

Summary: GAD67 is located at inhibitory synapses, chiefly in their presynaptic compartment.

Reason: The incoming broad synapse transfer is supported by the actual P48318 donor association with PMID:15479642, whose detailed GAD67 assay remains inaccessible. Independent original GAD67 presynaptic-marker and inhibitory-synapse measurements in PMID:23258346 and 24613359 establish the whole location. Retain the broad source term with explicit presynaptic interpretation; this does not imply postsynaptic receptor localization or recovery of every original donor panel.

Summary: The specific perinuclear localization transfer remains unverified.

Reason: The actual P48318 donor carries an IDA annotation to PMID:29960053, whose accessible abstract concerns Brinp1 disruption. That title does not show misattribution, but the relevant GAD67 protein staining and perinuclear compartment assignment were not recovered. Independent cytoplasmic GAD67 staining establishes a broader compartment, not this specific perinuclear enrichment. Retain uncertainty rather than replacing the precise donor observation with a general localization assumption.

Summary: The supported ancestral placement retains GAD67 at the presynaptic active zone.

Reason: The literal PTN002903810 assertion lies on the exact Q99259 lineage and is grounded in actual mouse and rat descendant annotations; no target-specific loss is demonstrated. Mouse donor records include PMID:12074840,18165320 and 23258346. Selected originals independently establish inhibitory GAD67-positive presynaptic terminals, while the exact active-zone antibody/panel detail in 12074840 was not accessible and must not be invented. Retain the considered PAINT/curatorial placement with that access limit; neither donor number nor a paper’s focus on another marker is a reason to reject it.

Summary: The mouse active-zone transfer is consistent with the independently curated ancestral placement.

Reason: The complete mouse donor capture contains actual IDA records from PMID:12074840,18165320 and 23258346, and the same location has a verified GAD-family IBD inherited by human Q99259. Selected original texts corroborate inhibitory presynaptic GAD67 protein, while the exact specialized active-zone spatial evidence is not independently reconstructed from those limited sections. Preserve the curatorial/phylogenetic location without asserting that SLC4A10 electron microscopy labelled GAD67 or that generic terminal fluorescence alone resolves active-zone membrane structure.

Summary: GAD67 is an experimentally used protein marker of inhibitory presynaptic sites.

Reason: The authentic original PMID:24613359 selected Results and figure captions measure GAD67-positive inhibitory presynaptic clusters and GAD67/gephyrin-labelled inhibitory synapses. The selected HEK293T co-culture assay uses rat hippocampal neurons; other cortical experiments include mice, and this distinction is preserved rather than silently equating one selected panel with the exact mouse donor annotation. Together with independent mouse GAD67 terminal staining, the conserved inhibitory-synapse location is supported. The synaptogenic perturbation belongs to calsyntenins, not GAD1.

Summary: The GABA-synthesis reaction explicitly includes vesicle-associated human GAD1.

Reason: The official R-HSA-888572 catalyst is the GAD1 dimer R-HSA-888570, containing two copies of Q99259 component R-HSA-888580. That component is explicitly assigned to GO:0061202. This resolves target membership rather than inferring a membrane pool from the reaction title. Retain the specific modeled transport-vesicle compartment as non-core; the enzyme also has broad cytoplasmic and presynaptic pools, and the model does not independently establish this targeting for every splice product.

Summary: The GAD2-labelled reaction also contains the human GAD1:GAD2 vesicle assembly.

Reason: Official R-HSA-888577 resolves through a decarboxylase-dimer set containing a GAD1:GAD2 heterodimer and the exact Q99259 GAD1 component R-HSA-888580. Thus its GAD2-oriented label does not demonstrate target absence. The GAD1 participant carries the specific GABA transport-vesicle membrane location. Retain this pool as non-core while distinguishing the curated assembly model from a newly read direct binding experiment or uniform isoform targeting.

Summary: The GABA-release event contains GAD1 through the loaded transport-vesicle complex.

Reason: The official event hierarchy resolves to the GAD1:GAD2 assembly and Q99259 component R-HSA-888580 with its specific transport-vesicle membrane assignment. This supports the source’s participant location and does not mean GAD1 catalyses membrane fusion or every event reaction. Retain the specialized vesicle pool as non-core, distinct from the unresolved event-level plasma-membrane annotation.

Summary: The vesicle-docking model explicitly includes the human GAD1 participant.

Reason: The official R-HSA-917744 hierarchy includes the docked-vesicle GAD1:GAD2 assembly and the exact Q99259 component R-HSA-888580. Its assigned transport-vesicle membrane location is retained as a specialized non-core pool. This is actual participant evidence, not inference from pathway membership alone, but it does not establish a direct GAD1 docking activity, membrane topology or equal targeting of all products.

Summary: GAD67 protein is present in inhibitory presynaptic terminals.

Reason: The literal mouse SynGO donor rows point to PMID:21584649, whose accessible abstract uses GFP at the Gad1 locus to define neuronal identity; its precise protein-localization assay remains uninspected. Independently, original GAD67 antibody Methods and terminal-marker Results in PMID:23258346 and 24613359 directly support the broader presynaptic location. Retain this conserved location while distinguishing independent corroboration from recovery of the cited donor assay. Donor anatomical extensions are not added to the human row.

Summary: GAD67 occupies the presynaptic side of GABAergic synapses.

Reason: The actual mouse SynGO records are retained with their original PMID:21584649 evidence and anatomical context; the inaccessible protein-level details are not reconstructed from a GFP cell-identity abstract. Independent original GAD67/VGAT terminal staining and GAD67/gephyrin inhibitory-synapse measurements support this location in GABAergic circuitry. This does not assign GAD1 postsynaptic receptor activity or transfer the donor brain-region extension to human.