GFI1B

Summary: GFI1B represses RNA polymerase II transcription.

Reason: The original study demonstrates Gfi1b-dependent repression using murine cDNA and mouse cells. Preserve that species limit rather than relabeling the experiment as direct human evidence. Human GFI1B consensus recognition, promoter occupancy and corepressor recruitment independently support the same biological function. The incoming IDA provenance remains unchanged.

Summary: GFI1B recognizes regulatory DNA sequences.

Reason: The asserted PTHR24393 tree places the exact human sequence below AN0/PTN002811836, with the literal DNA-recognition IBD and no retained negative record on its path. Human GFI1B sequence-selection and consensus-DNA binding experiments independently support recognition; the SELEX fragment establishes binding, while promoter occupancy and transcriptional work establish the regulatory context. The current UniProt PTHR23226 cross-reference does not replace the historical asserted book.

Summary: Sequence-specific transcriptional repression is the principal regulatory activity.

Reason: The NAS citation describes GATA1/Gfi1b repression complexes and cites prior DNA recognition; its own ChIP is promoter association rather than a purified DNA-binding assay. Human consensus-DNA binding and corepressor recruitment, together with the original repression experiments, justify the more informative repressor activity. Context-dependent activation is separately retained and is not denied by this refinement.

Proposed replacements: DNA-binding transcription repressor activity, RNA polymerase II-specific

Summary: Transcriptional activation is supported in restricted promoter contexts.

Reason: The actual activator IBD is at AN0/PTN002811836, ancestral to GFI1B; the mapped donors retain experimental activator annotations. Independent Gfi1b experiments show activation of a synthetic consensus-site promoter but repression of SOCS1. Those constructs are murine, even in human K562 recipients. Retain the phylogenetic assertion as a context-dependent capacity, without making constitutive activation or a particular endogenous human target the core mechanism.

Summary: The reported RUNX1T1 interaction supports corepressor binding.

Reason: The authentic abstract explicitly identifies ETO as a partner of Gfi-1B, although much of the detailed mechanistic summary concerns Gfi-1. Official Q06455 is RUNX1T1/ETO, a transcriptional corepressor. Refine the positive curator interaction to that established partner class. The inaccessible original leaves exact Gfi-1B constructs and assay details unverified; no purified-affinity, HDAC-binding or isoform-specific claim is added.

Proposed replacements: transcription corepressor binding

Summary: The NELL2 interaction is retained as evidence but generic binding is uninformative.

Reason: The exact GFI1B–NELL2 pair has positive Y2H and HEK293T pull-down records. NELL2 is a secreted ligand rather than a receptor: its class therefore does not support receptor binding by GFI1B. Receptor or receptor-ligand activity would additionally require signal transmission or a change in receptor activity, neither established here. Remove the generic term while retaining the positive association; this is not evidence that the pair is false.

Summary: Binding the microtubule regulator CPAP supports a specific partner-class term.

Reason: The exact CPAP/CENPJ pair has positive Y2H (EBI-952822) and HEK293T pull-down (EBI-946647) records from PMID:16713569. GO:0005519 describes binding a protein that regulates cytoskeletal reorganization; CPAP’s experimentally supported microtubule-growth function meets that definition. Retain this measured binding outside the core repression mechanism. It does not establish GFI1B control of microtubules, endogenous colocalization or physiological significance. Unlike receptor or ligand activity, this term requires no inferred signaling role for GFI1B. Historical assay constructs remain unmatched to individual natural products.

Proposed replacements: cytoskeletal regulatory protein binding

Summary: The JAG2 association does not define a specific GFI1B signaling activity.

Reason: Positive Y2H and HEK293T pull-down records support the exact GFI1B–JAG2 pair. The official record describes JAG2 as a putative Notch ligand. Binding that ligand does not establish GFI1B receptor binding, signal transmission or alteration of receptor activity. Remove the uninformative generic term while preserving the association and its construct limits; no negative interaction or physiological colocalization claim is made.

Summary: The inherited nuclear location agrees with GFI1B activity.

Reason: The exact human sequence descends from the nuclear IBD at AN0/PTN002811836. Independent human transcriptional complexes, promoter occupancy and the reviewed nuclear-location record support this compartment. Preserve the incoming is_active_in qualifier; the donor list and target self-support are not circularity.

Summary: Nuclear localization is supported despite limited access to the cited original.

Reason: The authentic source abstract explicitly places a portion of Gfi-1B in the nuclear matrix, which entails a nuclear pool. Other human GFI1B transcriptional and localization evidence supports the compartment, so the curator IDA is retained. Exact experiments in this original remain unavailable and are not reconstructed from its Gfi-1-focused title.

Summary: The UniProt nuclear-location mapping is supported.

Reason: The official reviewed GFI1B record contains the nucleus subcellular-location assertion SL-0191, matching the incoming electronic mapping. Human nuclear transcriptional work is consistent with it. This does not imply that all nuclear subcompartments or every natural product have been measured separately.

Summary: The curated nucleoplasmic localization is consistent with nuclear regulation.

Reason: Retain the HPA immunofluorescence IDA at its recorded resolution. GFI1B is a nuclear transcriptional regulator with human promoter-binding and nuclear-complex evidence. The individual HPA images have not been independently re-read here; no exclusive localization or isoform-specific distribution is inferred.

Summary: GFI1B participates in transcription-regulatory protein complexes.

Reason: The original mouse MEL study identifies a GATA1/Gfi1b complex and promoter association; it does not establish a Gfi1b–FOG1 complex. Human GFI1B/LSD1/CoREST co-immunoprecipitation provides separate complex evidence. These results support the generic transcription-regulator complex without treating every co-member pair as direct binding or attributing histone-demethylase catalysis to GFI1B.

Summary: The broad transcription-regulation term is covered by retained directional activities.

Reason: The actual AN0/PTN002811836 IBD and human GFI1B descendant are valid, and target self-evidence is legitimate. The annotation is true. Accepted repression and separately retained context-dependent positive regulation both have verified direct is_a paths to GO:0006357, making this undirected term redundant in the curated functional summary. No negative phylogenetic conclusion is implied.

Summary: The reported nuclear-matrix fraction is retained with its access limits.

Reason: The authentic abstract specifically reports that a portion of Gfi-1B associates with the nuclear matrix. This positive target statement supports retaining the curator observation. The full original could not be accessed, so fractionation, detergent resistance and product identity are not independently established. Matrix association is not made the organizing core function.

Summary: Nuclear-matrix association is supported as a secondary localization.

Reason: The incoming ARBA rule is not independently re-executed, but the authentic experimental source explicitly reports a Gfi-1B-associated nuclear-matrix fraction. Retain this limited localization without turning it into a structural matrix function or an exclusive nuclear compartment. Exact extraction conditions and natural-product mapping remain unresolved.

Summary: Positive transcriptional regulation is a context-dependent secondary capability.

Reason: This electronic process assertion follows the retained activator activity. Gfi1b activates an artificial consensus-site reporter in specific cell contexts while repressing a different promoter; the original uses murine cDNA. Retain the broad inferred positive process without asserting an endogenous human activated target, generalizing all transcriptional responses, or treating the reporter result as constitutive activation.

Summary: GFI1B binds the sequence-specific transcription factor GATA1 during repression.

Reason: Human GFI1B forms a complex with tagged mouse GATA1 in 293T cells, and GATA1 depletion removes human GFI1B recruitment to its own promoter in K562 cells (PMID:16177182). The original mouse MEL experiments also recover a GATA1/Gfi1b complex. Together these support transcription-factor binding as part of the repression mechanism. The incoming Q8IX07 identifier remains ZFPM1/FOG1: the cited MEL FOG1 immunoprecipitation did not recover Gfi1b. Accept the supported molecular-function claim while retaining the probable WITH/FROM defect for source review; this neither validates a positive FOG1 interaction nor silently substitutes GATA1 in the machine record.

Summary: Refine blood-cell developmental regulation to erythrocyte and megakaryocyte differentiation.

Reason: The actual AN29/PTN008616677 IBD supports inherited regulation of hemopoiesis without a phylogenetic failure. Target-specific evidence permits two more precise regulatory processes. In human UT-7 cells, p32-selective depletion impairs erythroid differentiation and p32 expression induces it without EPO (PMID:22399799); p37 overexpression also raises endogenous p32, so autonomous p37 sufficiency is unresolved. Human GFI1B mutations alter transcriptional activity and accompany abnormal megakaryocyte development and marker expression (PMID:23927492; PMID:24325358). The latter abstracts support a regulatory role with full experimental protocols uninspected. These unsigned refinements do not assign every natural product the same lineage role, infer a megakaryocyte-specific PAINT node, or propose new annotation rows.

Proposed replacements: regulation of erythrocyte differentiation regulation of megakaryocyte differentiation

Summary: The DNA-recognition annotation can be localized to a transcriptional regulatory role.

Reason: The actual SELEX workbook identifies a GFI1B extended DNA-binding domain matching canonical residues 148–330 and recovered methylated and unmethylated DNA motifs. It is an engineered 183-residue fragment, not a natural isoform. SELEX alone does not establish cis-regulatory occupancy; the independent PAINT regulatory-DNA assertion and human promoter/consensus-binding evidence supply that context. GO:0000978 is an actual is_a descendant of the incoming broad DNA-binding term.

Proposed replacements: RNA polymerase II cis-regulatory region sequence-specific DNA binding

Summary: The cited experiment has the opposite sign from inhibition of G1/S progression.

Reason: GO:2000134 decreases CDK activity to restrain the G1-to-S switch. In the inspected original, wild-type murine Gfi1b suppresses p21 and prevents IL6-induced G1 arrest in mouse M1 cells, contradicting the incoming TAS direction. The repression claim has independent human mechanistic support, whereas this conditional mouse cell-cycle result and human leukemia proliferation data do not establish the specific opposite human G1/S process. Retain the term/source query without adding GO:2000045.