GALK1

Summary: Human GALK1 expression produces galactokinase activity in the original comparison with GK2.

Reason: The authentic PMID:7542884 abstract explicitly reports expression of human GALK1 and GK2 cDNAs in COS cells and high galactokinase activity from GALK1. This is positive target-specific experimental support, corroborated by later purified-human-enzyme work. Full original Methods were not inspected, so equivalent protein expression or universal inactivity of the separate GK2 protein is not inferred.

Summary: Human GALK1 inherits experimentally grounded galactokinase activity from the inspected PAINT ancestral node.

Reason: The actual PTHR10457 tree places human P51570 at AN489/PTN002470832 beneath the activity IBD at AN0/PTN000048421. The incoming donor set matches the retained IBD, including the legitimate human self-donor; every donor identifier was resolved to the actual tree. The one activity-loss annotation is on another lineage. The human 392-residue sequence matches the MSA and independently measured activity in PMID:7542884 and PMID:12694189. There is no target-specific loss or divergence contradicting the ancestral placement.

Summary: Purified recombinant human GALK1 catalyzes ATP-dependent galactose phosphorylation.

Reason: PMID:12694189 expressed a sequence-checked human IMAGE3501788 coding sequence as an N-terminal His6 fusion in E.coli and measured ADP generation with a coupled assay. Active wild-type enzyme and substrate-dependent kinetics establish galactokinase activity. NADH is the reporter for coupled lactate dehydrogenase, not a demonstrated GALK1 redox cofactor; variant solubility/activity defects are not transferred to the normal human product.

Summary: The original human COS-cell experiment supports the specific galactokinase reaction.

Reason: PMID:7542884 directly attributes the increased measured galactokinase activity to human GALK1 cDNA expression. Preserve this IDA context separately from the source's EXP row; neither is a new assertion. Later recombinant-human kinetics independently confirm the reaction. The available source scope is the authentic abstract, not an invented full assay protocol.

Summary: The electronic galactokinase mapping agrees with direct human GALK1 biochemistry.

Reason: The literal EC/Rhea/InterPro/orthology support is preserved. Human GALK1 expressed in COS cells increases galactokinase activity (PMID:7542884), and purified recombinant human enzyme phosphorylates galactose using ATP (PMID:12694189; PMID:14596685). These direct findings independently establish the specific reaction, so the combined electronic mapping is biologically sound rather than accepted merely from its family label.

Summary: Galactokinase activity is established for normal GALK1, while this original Reactome source models defective variants.

Reason: Official R-HSA-5610026 is a FailedReaction with catalyst9631933 and mutant set5610011; it has no reaction output and explicitly points to normalReaction70355. Its activity label is not evidence that the disease variants retain normal turnover. The linked normal reaction70355→70354→70353 identifies P51570(1–392), and direct human experiments in PMID:7542884/12694189 establish the gene-level activity. Retain the original TAS source with this variant/model limitation instead of interpreting defective GALK1 as evidence against the normal enzyme.

Summary: The resolved HuRI interaction is retained as evidence, but generic protein binding is not an informative GALK1 function.

Reason: The authentic HuRI PSI dataset contains three assay records for P51570-1 with Q9NPJ4-1, including pooled, array and validated yeast two-hybrid contexts. These are positive observations of one human-ORF pair in yeast, not evidence of absence or a wrong-gene citation. The current GO:0005515 assertion supplies no specific biochemical role, and these data do not support an informative binding-function replacement. Remove the uninformative annotation under the generic-binding policy while preserving the exact pair, clone identifiers, assay methods and source suffixes in the evidence archive. The historical P51570-1 identifier is not converted into a newly declared natural isoform.

Summary: The original kinetic study supports ATP binding before galactose in the human enzyme.

Reason: PMID:12694189 measured galactose-1-phosphate inhibition patterns with purified human GALK1. Mixed inhibition with respect to ATP and uncompetitive inhibition with respect to galactose support an ordered ternary-complex mechanism with ATP binding first. This is mechanistic substrate-recognition evidence; neither Km alone nor use of an ATP-dependent assay is treated as a direct equilibrium binding constant.

Summary: ATP recognition is a demonstrated part of the GALK1 catalytic mechanism.

Reason: The InterPro-derived ATP-binding assertion is independently supported by human-enzyme product-inhibition kinetics indicating ATP-first substrate binding (PMID:12694189) and the human structure with Mg-AMPPNP (PMID:15590630). The structural ligand is an ATP analogue, not ATP turnover by itself. These complementary observations establish nucleotide recognition; they do not imply NADH binding by GALK1 from the coupled reporter assay.

Summary: Human GALK1 specifically recognizes galactose in its catalytic substrate pocket.

Reason: PMID:14596685 tested recombinant human substrate specificity and sugar-pocket mutations; its introductory structural figure is from bacterial L.lactis and is not relabeled as a human structure. The independently read human complex in PMID:15590630 directly contains alpha-D-galactose. Together these support galactose recognition. Robust in-vitro turnover of 2-deoxygalactose does not imply absolute specificity for galactose, and kinetic Km values are not equated with binding affinity.

Summary: The original fractionation experiment supports the more precise cytosolic location.

Reason: In PMID:8908517, newly translated human GALK1 remains proteinase-K accessible and in the supernatant rather than the microsomal pellet at both pH7.5 and pH11.5 (Results/Figure3 and Methods). This directly supports a soluble cytosolic setting within the tested system. GO:0005829 part_of GO:0005737 permits refinement of the original cytoplasm term. The same source's literal NOT-membrane annotation remains separately preserved.

Proposed replacements: cytosol

Summary: The broad cytoplasm mapping can be refined to the experimentally supported cytosol.

Reason: Human GALK1 remains soluble and proteinase-K accessible with microsomal membranes in the original expression/fractionation experiment (PMID:8908517), while the normal P51570 Reactome entity is in cytosol. GO:0005829 part_of GO:0005737 is directly verified. Cytosol is therefore a supported compartment refinement of this generic cytoplasm mapping; no organelle location or universal exclusion of all peripheral associations is inferred.

Proposed replacements: cytosol

Summary: The actual PAINT cytosol placement agrees with human soluble-enzyme evidence.

Reason: The cytosol IBD at AN0/PTN000048421 propagates to exact human AN489/PTN002470832, with retained donor evidence and no contradictory target loss. Human GALK1 is soluble and fails to insert into or pellet with microsomes in PMID:8908517; the normal Reactome physical entity is also cytosolic. The original is_active_in relationship is retained. Potential context-specific vesicle-fraction occurrence does not contradict a nonexclusive cytosolic core location.

Summary: The disease model locates its GALK1 mutant set in cytosol, consistent with independent normal-protein evidence.

Reason: Official FailedReaction5610026 uses catalyst9631933 and the GALK1 mutant set5610011, whose compartment is explicitly GO:0005829. This disease model is not a normal-protein localization assay, but the normal linked reaction70355 uses P51570 entity70353 in the same compartment, and PMID:8908517 supplies independent human soluble-protein evidence. Retain the valid gene-level cytosol claim with this source-model boundary.

Summary: The normal GALK1 reaction explicitly places P51570 in cytosol.

Reason: Official reaction70355→catalyst70354→physical entity70353 identifies human UniProt P51570, coordinates1–392, in GO:0005829. The reaction converts galactose and ATP to galactose1-phosphate and ADP. This curated location is coherent with the independently read human microsomal-exclusion experiment; unrelated pathway members and mutant models are not substituted for the target.

Summary: The valid ancestral galactose-metabolism assertion can be refined to GALK1's demonstrated Leloir-pathway step.

Reason: The actual metabolic-process IBD is at AN0/PTN000048421, ancestral to the exact human AN489 sequence; donor evidence and descent are intact. Human GALK1 catalyzes the first committed phosphorylation step of galactose catabolism, rather than the reversible UDP-sugar interconversion performed by GALE. Official GO paths establish GO:0033499 is_a GO:0019388 is_a GO:0006012. The replacement adds supported biological precision without treating the broad ancestral assertion as erroneous or downgrading its self-donor grounding.

Proposed replacements: beta-D-galactose catabolic process via UDP-galactose, Leloir pathway

Summary: The electronic galactose-metabolism assertion can be refined to the directly supported Leloir pathway.

Reason: The original EC/UniPathway/orthology context is preserved. Human GALK1 biochemistry independently establishes the committed galactose-phosphorylation step of the Leloir pathway. GO:0033499 reaches GO:0006012 through two verified is_a edges via galactose catabolism, so the narrower process is a genuine descendant rather than a semantic substitute.

Proposed replacements: beta-D-galactose catabolic process via UDP-galactose, Leloir pathway

Summary: The human variant-linked metabolic process is more precisely described by the established Leloir-pathway role.

Reason: The authentic PMID:15024738 abstract identifies human V32M and R239Q variants and altered activity after mutant-cDNA expression in E.coli; the full original Methods were not inspected. This is compatible with the core process and is not rejected because it is variant-focused. Independent wild-type human kinetics and the normal Reactome reaction establish GALK1's Leloir step. The verified GO:0033499→GO:0019388→GO:0006012 is_a path permits a precise process replacement without claiming that the abstract alone assayed complete pathway flux.

Proposed replacements: beta-D-galactose catabolic process via UDP-galactose, Leloir pathway

Summary: A single high-throughput detection in an NK-cell membrane preparation does not establish that the soluble cytosolic galactokinase is a membrane protein, and the targeted experiment on the same gene product contradicts the positive claim.

Reason: PMID:19946888 is a bulk membrane-proteome survey of YTS cells (abstract only in the cache). Its own abstract states that only about 40% of the 1843 identified proteins were predicted to be membrane proteins and that the remainder were proteins predicted to be transiently associated with membranes, so presence in the preparation is not evidence of membrane residence. GALK1 has no transmembrane segment or lipid anchor, the human structure and kinetics describe a soluble enzyme, all curated Reactome entities place it in cytosol, and the targeted in vitro translation experiment on human GALK1 (PMID:8908517) concluded that the protein is cytosolic and not associated with the endoplasmic reticulum membrane, which UniProt itself recorded as a NOT located_in membrane annotation. The generic positive membrane claim is therefore contradicted rather than merely unverified. Removal does not deny that some GALK1 co-purified with the YTS membrane fraction; it records that this does not make membrane a location of the gene product. The GALK1-specific peptide table was not inspected.

Summary: The original experiment supports the retained NOT-membrane assertion in its tested microsomal context.

Reason: This original is NOT|located_in, not a positive membrane annotation. In PMID:8908517, human GALK1 translated in rabbit reticulocyte lysate remained proteinase-K accessible in the presence of canine pancreatic microsomes and stayed in the supernatant after pelleting at both neutral and carbonate pH (Results/Figure3 and expression Methods). Retain negated:true and the original source. This supports absence of insertion/association in that assay; it does not establish that no partner-dependent peripheral association can ever occur in YTS or another cell type.

Summary: The broad kinase assertion can be refined to the verified galactokinase activity.

Reason: Human biochemical experiments establish ATP-dependent phosphorylation of galactose at C1. Official GO paths verify GO:0004335 is_a GO:0019200 is_a GO:0016301. Thus galactokinase activity is a supported descendant of the original broad kinase term, not a lateral replacement or a newly invented function.

Proposed replacements: galactokinase activity

Summary: The alcohol-acceptor phosphotransferase mapping can be refined to galactokinase activity.

Reason: The C1 hydroxyl of galactose accepts phosphate from ATP in the directly measured GALK1 reaction. Official GO records verify GO:0004335 is_a GO:0016773. Replace the broad transferase term with this experimentally established descendant while retaining the original InterPro evidence and context.

Proposed replacements: galactokinase activity

Summary: The transferred Leloir-pathway assertion is independently supported by human GALK1 experiments.

Reason: Preserve the exact mouse/Ensembl source identifiers and current official GO:0033499 label. Human GALK1 performs the initial galactose phosphorylation required for the canonical Leloir sequence, as shown by recombinant-human biochemistry and the normal target-specific Reactome reaction. The current label differs from the inherited wording, but the GO identifier and original annotation context remain intact; no mouse-specific kinetic or assembly property is assumed.

Summary: The curated pathway explicitly includes GALK1's galactose-phosphorylation reaction.

Reason: Official R-HSA-70370 contains the normal reaction70355, whose catalyst70354 leads to P51570 physical entity70353 in cytosol. GALK1 catalyzes a chemical step of this pathway, not merely a substrate consumed by another enzyme. The overall pathway also contains other enzymes and branches; those functions are not transferred to GALK1.

Summary: Carbohydrate phosphorylation is a true process directly performed by GALK1.

Reason: GALK1 transfers phosphate from ATP to galactose, satisfying the official definition of introducing a phosphate group into a carbohydrate. The specific Leloir-pathway term is already represented, but the current ontology has no is_a/part_of path from GO:0033499 to GO:0046835. Therefore that term is not presented as a hierarchical replacement for this valid assertion. The reaction evidence supports retention of the original phosphorylation process.

Summary: Original urinary-exosome proteomics identifies GALK1 as a limited non-core fraction-associated occurrence.

Reason: PMID:19056867 Table1 explicitly names GALK1, and the authentic expanded NHLBI table identifies GALK1/NP_000145 with peptide count1 and reference2 corresponding to this paper. Selected original Methods document differential centrifugation, LC-MS/MS and single-peptide/ambiguity filtering. These are original experimental data, not circular use of the table's imported GO labels. No target peptide sequence or independent encapsulation/topology assay was recovered. Retain limited urinary-vesicle-fraction occurrence as non-core, with the exact part_of UBERON:0001088 atom, without claiming a secretion mechanism or vesicle function.