Ang2

Summary: Phylogenetic RNase transfer matches direct mouse biochemical evidence.

Reason: Mouse Ang2 has direct RNA/tRNA cleavage evidence.

Summary: Secreted extracellular localization is plausible, but it is a localization note rather than a core function.

Reason: Signal peptide and secreted RNase family context support extracellular localization.

Summary: Angiogenin-subfamily antibacterial transfer is defensible for mouse Ang2 but is not a core function.

Reason: The IBD behind this IBA sits at PANTHER node PTN000153659 and is seeded by human angiogenin (UniProtKB:P03950) together with MGI:MGI:2656551, i.e. angiogenin-subfamily members that are among the closest relatives of Ang2 rather than out-of-family donors, and mouse Ang2 lies inside that clade. Ang2/Angrp remains a secreted, catalytically active RNase (PMID:8633065, PMID:23170778) and no target-specific evidence of loss of antimicrobial function exists, so paralog proximity alone is triage evidence rather than grounds for removal. Kept as non-core because there is no direct mouse Ang2 antibacterial experiment; the same node's rows on human ANG are also kept as non-core.

Summary: Angiogenin-subfamily innate-immune transfer is defensible for mouse Ang2 but is not a core function.

Reason: The IBD for this term at PANTHER node PTN000153659 is seeded by human angiogenin (UniProtKB:P03950), an angiogenin-subfamily member rather than an out-of-family donor, and mouse Ang2 descends from that node. Mouse Ang2/Angrp is a secreted RNase that retains full catalytic activity (PMID:8633065, PMID:23170778), and there is no evidence specific to the mouse protein that it has lost an immune effector role, so the family inference cannot be refuted on paralog-proximity grounds alone. Kept as non-core: no direct mouse Ang2 immune experiment supports it, and this is a broad process term rather than a core Ang2 function.

Summary: Gram-positive specificity comes from the antimicrobial-RNase branch of the family, not from the angiogenin subfamily that contains Ang2.

Reason: This row propagates from the broader node PANTHER:PTN008517922. Of the ten IBD seeds, seven could be identified here, by two different routes. Six are UniProt accessions and so are direct keys in PTHR11437-entries.csv. The remaining one, MGI:MGI:1925666, is not a key in that accession-keyed file at all; it was corroborated through the UniProt cross-reference for this MGI id, and it is that accession (Q9D244) which entries.csv places. The seven split as follows. Five sit in the antimicrobial-RNase branch of the RNase A family: UniProtKB:P12724 RNASE3 (SF3), UniProtKB:Q93091 RNASE6 (SF4), UniProtKB:Q9H1E1 RNASE7 (SF31), UniProtKB:Q8TDE3 RNASE8 (SF25) and MGI:MGI:1925666, mouse Rnase6 (Q9D244, SF4). One is angiogenin itself, UniProtKB:P03950 (SF60). The seventh, chicken UniProtKB:P30374, is classified SF10 ANGIOGENIN-RELATED and so belongs to neither group. The remaining three are zebrafish ZFIN ids, cited by identifier with no subfamily asserted: entries.csv is accession-keyed and carries no ZFIN ids, so nothing local resolves them. Generic secreted-RNase antimicrobial biology may transfer to Ang2, so the row is not simply wrong, but the pathogen-class specificity of the term is anchored in experiments on the RNASE3/6/7/8 branch and is not supported for the angiogenin-subfamily mouse protein. That the branch experiments carry the pathogen-class specificity is asserted from external knowledge of RNase A family biology and is NOT corroborable here: no cached Ang2 record mentions RNASE3/6/7/8 or eosinophil biology, the deep-research file's only antimicrobial passage concerns Ang1/Ang4 in colitis and states that it does not address Ang2, and Ang2-bioinformatics/RESULTS.md raises antimicrobial terms only to note that their human source is itself inferred. It is not the only ground for the verdict, however, and the other is local: as the UniProtKB:P03950 source entity below records, the one SF60 seed this review could identify among the node's seeds for this term carries its own GO:0050830 as an IBA on this same node, with no experimental row behind it. That covers P03950 alone rather than the whole transfer: this row's GOA WITH/FROM also carries MGI:MGI:88022, likewise SF60, whose equivalent negative cannot be checked locally because there is no genes/mouse/Ang directory. As far as it reaches, it supports the same over-annotation call together with the RESULTS.md observation just cited, and without recourse to RNase A family knowledge. The row is treated as an over-annotation rather than removed.

Summary: Angiogenin-subfamily antimicrobial transfer is defensible for mouse Ang2 but is not a core function.

Reason: As for the sibling immune terms, the IBD at PANTHER node PTN000153659 is seeded by human angiogenin (UniProtKB:P03950) within the angiogenin subfamily that contains mouse Ang2, and Ang2/Angrp retains secretion and full ribonucleolytic activity (PMID:8633065, PMID:23170778). No mouse experiment demonstrates or excludes an antimicrobial role, so the transfer is retained as non-core rather than removed on the strength of paralog proximity.

Summary: Angiogenesis is directly refuted for the mouse Ang2/Angrp protein.

Reason: REMOVE survives the checklist because the objection is target-specific, not merely that the donors are paralogs. The cached PMID:8633065 states verbatim that mouse Ang is potently angiogenic but Angrp is not, even at relatively high doses, and attributes this to an inability to bind cellular receptors (poor conservation of the Ang receptor recognition segment 58-69) rather than to loss of catalysis, since Angrp is if anything the more active ribonuclease. PMID:23170778 independently records that mAng-2 is not angiogenic. The angiogenin-subfamily donors genuinely are angiogenic, so the source annotations are sound and only the transfer to Ang2 fails.

Summary: Generic nucleic acid binding is less informative than the specific RNase activity.

Reason: Prefer RNA nuclease activity.

Summary: DNA binding is inferred from canonical angiogenin biology, not demonstrated for mouse Ang2.

Reason: No direct mouse evidence supports DNA binding for Ang2.

Summary: Generic nuclease parent term; RNA nuclease activity is the informative term.

Reason: Specific RNA nuclease terms better capture the supported mouse Ang2/Angrp activity.

Proposed replacements: RNA nuclease activity tRNA-specific ribonuclease activity

Summary: Generic endonuclease activity is broader than the supported RNase activity.

Reason: Specific RNA nuclease terms better capture the supported mouse Ang2/Angrp activity.

Proposed replacements: RNA nuclease activity tRNA-specific ribonuclease activity

Summary: Electronic RNase prediction matches direct mouse biochemical evidence.

Reason: Consistent with direct mouse RNase assays.

Summary: Extracellular localization is plausible for a secreted Ang2 protein and is used as the core-function location.

Reason: Secreted extracellular-region localization is consistent with the core RNase function.

Summary: Nucleolar localization is imported from canonical angiogenin biology and is not shown for mouse Ang2.

Reason: This term is an automated import of canonical angiogenin biology onto Angrp, a divergent paralog. PMID:8633065 finds Angrp is not angiogenic and implicates an inability to bind cellular receptors, with poor conservation of the Ang receptor recognition sequence 58-69 — while its ribonucleolytic activity toward tRNA is somewhat greater than Ang's, which is why this review ACCEPTs GO:0004540. Retained RNase activity does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the automated transfer is unsafe rather than merely unverified.

Summary: Generic hydrolase parent term; RNA nuclease activity is the informative term.

Reason: Specific RNA nuclease terms better capture the supported mouse Ang2/Angrp activity.

Proposed replacements: RNA nuclease activity tRNA-specific ribonuclease activity

Summary: Cytoplasmic-vesicle localization is a speculative automated transfer, not demonstrated for mouse Ang2.

Reason: This term is an automated import of canonical angiogenin biology onto Angrp, a divergent paralog. PMID:8633065 finds Angrp is not angiogenic and implicates an inability to bind cellular receptors, with poor conservation of the Ang receptor recognition sequence 58-69 — while its ribonucleolytic activity toward tRNA is somewhat greater than Ang's, which is why this review ACCEPTs GO:0004540. Retained RNase activity does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the automated transfer is unsafe rather than merely unverified.

Summary: Metal binding is supported by the mouse structural study, but it is secondary to the core RNase activity.

Reason: Mouse Ang2 binds Zn2+ and shows metal-sensitive RNase activity.

Summary: Transferred from human ANG via GO_REF:0000119, but the stronger is_active_in qualifier is not demonstrated for mouse Ang2 and is redundant with broader secretion annotations.

Reason: Extracellular localization is plausible, but the qualifier is stronger than the direct mouse evidence.

Summary: Transferred from human ANG via GO_REF:0000119, but mouse Ang2/Angrp is explicitly non-angiogenic.

Reason: PMID:8633065 contradicts the transfer.

Summary: Transferred from human ANG hypoxia biology, not shown for mouse Ang2.

Reason: Divergent paralog with no direct mouse evidence for this process.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source no longer carries this term.

Reason: Stale ISO transfer.

Summary: Transferred endothelial-proliferation biology from human ANG, but mouse Ang2 is non-angiogenic.

Reason: PMID:8633065 shows Angrp does not share canonical angiogenic biology.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source no longer carries this term.

Reason: Stale ISO transfer.

Summary: Transferred from human ANG actin-binding studies, but mouse Ang2 receptor/interface biology diverges from canonical angiogenin.

Reason: No direct mouse evidence and paralog divergence makes transfer unsafe.

Summary: Transferred from human ANG via GO_REF:0000119 and independently supported by mouse Ang2 biochemical assays.

Reason: Direct mouse RNase evidence exists.

Summary: Transferred from human ANG via GO_REF:0000119 and directly supported by mouse Ang2 tRNA cleavage assays.

Reason: Direct mouse tRNA cleavage evidence exists.

Summary: Transferred from human ANG receptor-binding studies, but mouse Ang2 lacks canonical receptor-mediated angiogenic biology.

Reason: PMID:8633065 implicates deficient receptor binding in Angrp.

Summary: Human ANG source supports copper binding, but mouse Ang2 data support only broader divalent-metal binding.

Reason: Mouse structural data support metal ion binding, not copper specificity.

Proposed replacements: metal ion binding

Summary: Transferred basement-membrane association from human ANG angiogenic biology, not shown for mouse Ang2.

Reason: Non-angiogenic paralog; no direct mouse evidence.

Summary: Transferred extracellular-space location from human ANG; plausible but redundant with retained non-ISO secretion annotations.

Reason: Extracellular localization is plausible, but it remains non-core and redundant.

Summary: Transferred from human ANG nuclear-activity studies, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred from human ANG nuclear-localization studies, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source no longer carries this term.

Reason: Stale ISO transfer.

Summary: Transferred nucleolar localization from human ANG, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred cytoplasmic activity from human ANG intracellular stress studies, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source no longer carries this term.

Reason: Stale ISO transfer.

Summary: Transferred signal transduction from human ANG, but mouse Ang2 lacks canonical angiogenic signaling evidence.

Reason: Non-angiogenic paralog with no direct mouse support.

Summary: Transferred heparin binding from human ANG, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred rRNA-transcription biology from human ANG, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred hormone-response biology from human ANG, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred stress-granule localization from human ANG stress-response studies, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source no longer carries this term.

Reason: Stale ISO transfer.

Summary: Transferred cell-migration biology from human ANG angiogenic studies, but mouse Ang2 is non-angiogenic.

Reason: No direct mouse evidence and canonical angiogenic transfer is unsafe.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source is itself inferred-only (IBA).

Reason: Do not stack inference on inference.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source no longer carries this term.

Reason: Stale ISO transfer.

Summary: Transferred generic signaling from human ANG, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred endocytic-vesicle localization from human ANG internalization studies, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred translational stress-response biology from human ANG, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred angiogenin-PRI complex membership from human ANG, not demonstrated for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source no longer carries this term.

Reason: Stale ISO transfer.

Summary: Transferred stress-granule assembly from human ANG, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source no longer carries this term.

Reason: Stale ISO transfer.

Summary: Transferred peptide-binding annotation from human ANG receptor/actin studies, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred positive regulation of phosphorylation from human ANG signaling studies, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred protein homodimerization from human ANG, but mouse Ang2 dimerization is not established as a physiological activity.

Reason: Structural observations are insufficient for this GO claim.

Summary: Transferred ribosome binding from human ANG stress-response studies, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source is itself inferred-only (IBA).

Reason: Do not stack inference on inference.

Summary: Transferred receptor-ligand activity from human ANG, but mouse Ang2 lacks canonical receptor-mediated angiogenic signaling evidence.

Reason: Non-angiogenic paralog with no direct mouse evidence.

Summary: Transferred smooth-muscle proliferation phenotype from human ANG, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Transferred positive regulation of protein secretion from human ANG, but mouse Ang2 lacks direct evidence for this process.

Reason: Source support is not sufficient to justify propagation to Ang2.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source is itself inferred-only (IBA).

Reason: Do not stack inference on inference.

Summary: Transferred from human ANG via GO_REF:0000119, but the current human source is itself inferred-only (IBA).

Reason: Do not stack inference on inference.

Summary: Transferred hematopoietic stem-cell proliferation from human ANG, not shown for mouse Ang2.

Reason: Angrp is a divergent paralog of angiogenin. PMID:8633065 compares the two directly and reports three results, not one: Angrp is not angiogenic; this is not a catalytic deficiency, since Angrp has somewhat greater ribonucleolytic activity toward tRNA and dinucleotide substrates than Ang, which is why this review ACCEPTs GO:0004540; and an inability to bind cellular receptors is implicated, with poor conservation of the Ang receptor recognition sequence 58-69. Retained RNase activity therefore does not carry ANG's receptor-dependent cellular biology across, and nothing in mouse Angrp supports this term independently, so the blanket GO_REF:0000119 transfer from human ANG (UniProtKB:P03950) is unsafe. Absence of mouse data alone would not justify removing an ISO, since that is what ISO asserts; the demonstrated divergence is the argument.

Summary: Cannot be adjudicated here: PMID:25313067 is cached abstract-only. Prior observation, not a verdict: Misannotation from AngII signaling, not Ang2/angiogenin-related protein.

Reason: PMID:25313067 is cached abstract-only, so the curator's experimental evidence cannot be inspected here; per CLAUDE.md an experimental annotation is not overruled on that basis. Recorded observation, not a verdict: The paper studies angiotensin II signaling through KLHL3 and WNK4.

Summary: Cannot be adjudicated here: PMID:28096417 is cached abstract-only. Prior observation, not a verdict: Misannotation from AngII signaling, not Ang2/angiogenin-related protein.

Reason: PMID:28096417 is cached abstract-only, so the curator's experimental evidence cannot be inspected here; per CLAUDE.md an experimental annotation is not overruled on that basis. Recorded observation, not a verdict: The paper studies WNK4 phosphorylation downstream of AngII and vasopressin.

Summary: Cannot be adjudicated here: PMID:25313067 is cached abstract-only. Prior observation, not a verdict: Misannotation from AngII signaling, not Ang2/angiogenin-related protein.

Reason: PMID:25313067 is cached abstract-only, so the curator's experimental evidence cannot be inspected here; per CLAUDE.md an experimental annotation is not overruled on that basis. Recorded observation, not a verdict: The paper is about angiotensin II regulation of KLHL3 and WNK4, not the Ang2 gene product.

Summary: Cannot be adjudicated here: PMID:28096417 is cached abstract-only. Prior observation, not a verdict: Misannotation from AngII signaling, not Ang2/angiogenin-related protein.

Reason: PMID:28096417 is cached abstract-only, so the curator's experimental evidence cannot be inspected here; per CLAUDE.md an experimental annotation is not overruled on that basis. Recorded observation, not a verdict: The paper is about WNK4 phosphorylation downstream of AngII and vasopressin, not Ang2.

Summary: Cannot be adjudicated here: PMID:25313067 is cached abstract-only. Prior observation, not a verdict: Misannotation from AngII signaling, not Ang2/angiogenin-related protein.

Reason: PMID:25313067 is cached abstract-only, so the curator's experimental evidence cannot be inspected here; per CLAUDE.md an experimental annotation is not overruled on that basis. Recorded observation, not a verdict: The paper concerns angiotensin II as a hormone stimulus, not the Ang2 gene product.

Summary: Cannot be adjudicated here: PMID:25313067 is cached abstract-only. Prior observation, not a verdict: Misannotation from AngII/WNK4 physiology, not Ang2/angiogenin-related protein.

Reason: PMID:25313067 is cached abstract-only, so the curator's experimental evidence cannot be inspected here; per CLAUDE.md an experimental annotation is not overruled on that basis. Recorded observation, not a verdict: This physiology belongs to the angiotensin II pathway, not the Ang2 gene.

Summary: Misannotation from a ZNF418 cardiac-hypertrophy study that uses Ang II as a stimulus.

Reason: The paper is about ZNF418, not Ang2/angiogenin-related protein.

Summary: Misannotation from a ZNF418 cardiac-hypertrophy study that uses Ang II as a stimulus.

Reason: The paper is about ZNF418, not Ang2/angiogenin-related protein.

Summary: Misannotation from a ZNF418 cardiac-hypertrophy study that uses Ang II as a stimulus.

Reason: The paper is about ZNF418, not Ang2/angiogenin-related protein.

Summary: Direct biochemical evidence for mouse Ang2 RNase activity.

Reason: On-target mouse structural and enzymatic study.

Summary: Cannot be adjudicated here: PMID:17670746 is cached abstract-only. Prior observation, not a verdict: Misannotation from an angiopoietin-2 paper, not Ang2/angiogenin-related protein.

Reason: PMID:17670746 is cached abstract-only, so the curator's experimental evidence cannot be inspected here; per CLAUDE.md an experimental annotation is not overruled on that basis. Recorded observation, not a verdict: The paper studies angiopoietin-2 (Ang-2), not mouse Ang2/Angrp.