AFF1 (also called AF4, and the founding member of the AF4/FMR2 family) is a large nuclear protein, about three-quarters of whose 1210 residues are intrinsically disordered, that works as a scaffold rather than as an enzyme or a sequence-specific DNA-binding factor. It is the subunit that holds together one branch of the super elongation complex: it is constitutively bound to the CDK9-cyclin T kinase P-TEFb and is present in every major P-TEFb assembly, and a short motif in its middle folds onto the ANC1-homology domain of AF9 with sub-nanomolar affinity, tying P-TEFb to the AF9/ENL and ELL/EAF modules on chromatin. Through that complex AFF1 promotes the release of promoter-paused RNA polymerase II into productive elongation, which depends on P-TEFb phosphorylating the polymerase C-terminal domain. Its own abundance and modification state make it a control point rather than a passive strut: p300 acetylation loosens the complex and shuts transcription down soon after DNA damage, while PARP1-dependent poly(ADP-ribosyl)ation protects AFF1 from Siah1-directed ubiquitination and supports transcriptional restart, and cells short of AFF1 repair DNA and survive genotoxic stress poorly. Its gene-specific outputs include the Wnt antagonist DKK1, through which AFF1 restrains osteoblast differentiation in mesenchymal stromal cells - the opposite of what its closest paralogue AFF4 does there. AFF1 is also the most frequent partner fused to KMT2A/MLL by the t(4;11)(q21;q23) translocation of infant acute lymphoblastic leukaemia; the resulting chimera is a distinct oncoprotein, and its properties are not simply those of the wild-type protein.
Definition: The resumption of RNA polymerase II transcription following the transcriptional shutdown that DNA damage induces, once lesions have been repaired or bypassed.
Justification: GO has no term for this process. Verified on a working endpoint rather than inferred from an empty search: a control query for "super elongation complex" returns GO:0032783 first, five phrasings of the restart concept return nothing relevant, and none of the 96 is_a/part_of descendants of GO:0006974 has "recover", "restart" or "resum" in its name. The consequence is visible in this gene's record: the one process that PMID:41062835 and PMID:31611376 jointly describe - AFF1 acetylation shutting SEC down early after genotoxic stress, AFF1 PARylation stabilising it for recovery - has to be expressed as an unrelated pair of terms, GO:0006974 plus GO:0032968, which loses the fact that AFF1 is the switch for both halves. The field treats recovery as a distinct, separately assayed phase of the damage response, so a child of GO:0006974 would let it be annotated as one process.
Parent term: DNA damage response
Supporting Evidence:
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0006355 regulation of DNA-templated transcription | IBA GO_REF:0000033 | ACCEPT | Summary: Accepted as a correct family-level statement. All four protein donors (mouse Aff1, mouse Aff2, mouse Aff3, Drosophila lilli) carry their own experimental annotations in this subtree, so the chain is not empty of experimental evidence and no source-weakness objection survives measurement. The unsigned parent is the right level here rather than lazy curation: the family is heterogeneous in DIRECTION - human AFF1 activates DKK1 while mouse Aff2 carries negative regulation of gene expression by IMP - so a signed child would be false for part of the clade. Reason: AFF1 regulates transcription as the scaffolding subunit of the super elongation complex, and the gene independently holds the specific, polymerase-resolved descendants of this term from human experiments. This row is the family-level generalisation of that, and it is a verified ancestor of GO:0032786 which the human IMP rows already assert, so it adds nothing false. One of the five WITH/FROM tokens is the target itself, which records a PAINT curator judging transcriptional regulation to be AFF1's core function. Propagation Review Root cause: NO FAILURE CORE Sources checked: FB:FBgn0041111 · Drosophila lilli (Q9VQI9) - the single fly family member, co-orthologue of all four vertebrate AFFs SUPPORTS TRANSFER the single fly family member, with its own experimental annotation in every subtree it is cited for MGI:MGI:106927 · mouse Aff3 (P51827) - paralog SUPPORTS TRANSFER a paralog rather than the ortholog, which is legitimate for IBA but means no ortholog-strength inference rests on this token; it carries its own IDA and IMP rows MGI:MGI:1100819 · mouse Aff1 (O88573) - the 1:1 ortholog SUPPORTS TRANSFER the true ortholog, and it carries its own IDA to positive regulation of DNA-templated transcription (PMID:9365243) PANTHER:PTN000829417 · PANTHER family node (AF4/FMR2 family; 79 recipient gene products, human reach = AFF1/AFF2/AFF3/AFF4) SUPPORTS TRANSFER an internal PANTHER tree node, not a protein; its human reach is exactly the four AFF paralogues, all of which belong to SEC-family complexes UniProtKB:P51825 · human AFF1 itself (self-referential IBD seed) SUPPORTS TRANSFER Self-reference: the target is its own IBD seed, which is expected rather than circular -- its own IMP annotation to GO:0032786 (positive regulation of DNA-templated transcription, elongation), a descendant of this term, is one of the descendant evidences behind the IBD. The IBA then asserts the additional claim that the function is inherited rather than lineage-specific. Supporting Evidence: PMID:22547686 We previously purified the AFF1- and AFF4-containing super elongation complex (SEC) as a major regulator of development and cancer pathogenesis. PMID:28955517 we performed an anti-AFF1 ChIP assay, which demonstrated that AFF1 bound to the promoter region of DKK1 |
| GO:0003712 transcription coregulator activity | IBA GO_REF:0000033 | ACCEPT | Summary: Accepted as the family's molecular function. The donor, Drosophila lilli, holds this exact term by IMP (PMID:11171404), so this is a transfer from an experimentally annotated family member rather than a family-level guess. The term's own definition permits action "either on its own or as part of a complex", which is what AFF1 does: it modulates transcription of specific gene sets from within SEC. Reason: AFF1 has no catalytic domain and no sequence-specific DNA-binding domain - 901 of its 1210 residues (74.5%) carry a Disordered feature, computed from the UniProt feature table rather than estimated - so a coregulator activity is exactly the right class of molecular function for it. Human evidence: it occupies the DKK1 promoter and sets DKK1 output in both directions in mesenchymal stromal cells. Recorded in core_functions under contributes_to_molecular_function, because the coregulatory output is delivered by the assembled complex. Propagation Review Root cause: NO FAILURE CORE Sources checked: FB:FBgn0041111 · Drosophila lilli (Q9VQI9) - the single fly family member, co-orthologue of all four vertebrate AFFs SUPPORTS TRANSFER the single fly family member, with its own experimental annotation in every subtree it is cited for PANTHER:PTN000829417 · PANTHER family node (AF4/FMR2 family; 79 recipient gene products, human reach = AFF1/AFF2/AFF3/AFF4) SUPPORTS TRANSFER an internal PANTHER tree node, not a protein; its human reach is exactly the four AFF paralogues, all of which belong to SEC-family complexes Supporting Evidence: PMID:28955517 we performed an anti-AFF1 ChIP assay, which demonstrated that AFF1 bound to the promoter region of DKK1 PMID:31611376 we show that site-specific acetylation of super elongation complex (SEC) subunit AFF1 by p300 reduces its interaction with other SEC components and impairs P-TEFb-mediated C-terminal domain phosphorylation of RNA polymerase II both in vitro and in vivo |
| GO:0006354 DNA-templated transcription elongation | IBA GO_REF:0000033 | MODIFY | Summary: The essence is right and the polymerase is under-specified. This row is self-referential - the WITH/FROM cites AFF1 itself alongside the node - so it records a PAINT curator judging transcription elongation to be AFF1's core function, which this review agrees with. But the same node asserts, of the same 79 recipient gene products, GO:0032783 super elongation complex, a term whose definition is explicitly about "RNA polymerase II transcription elongation". A polymerase-agnostic process term and a Pol II-specific complex term cannot both be maximally precise. Reason: Every characterised SEC substrate is RNA polymerase II: SEC-family complexes are assayed as Pol II CTD kinases, the p300 study measures Pol II CTD phosphorylation specifically, and AFF1 already holds the Pol II-resolved regulatory term GO:0032968 by IMP from human knockdown. There is no evidence of AFF1 acting on RNA polymerase I or III. GO:0006368 was verified by fetching its is_a/part_of closure to be a genuine descendant of GO:0006354, so this is a one-step refinement that asserts nothing new about the polymerase beyond what the gene's other rows already say. Propagation Review Root cause: NO FAILURE CORE Failure modes: GRANULARITY MISMATCH Sources checked: PANTHER:PTN000829417 · PANTHER family node (AF4/FMR2 family; 79 recipient gene products, human reach = AFF1/AFF2/AFF3/AFF4) SUPPORTS TRANSFER an internal PANTHER tree node, not a protein; its human reach is exactly the four AFF paralogues, all of which belong to SEC-family complexes UniProtKB:P51825 · human AFF1 itself (self-referential IBD seed) SUPPORTS TRANSFER Self-reference: the target is its own IBD seed, which is expected rather than circular -- its own EXP annotation to this term is one of the descendant evidences behind the IBD. The IBA then asserts the additional claim that the function is inherited rather than lineage-specific. Proposed replacements: transcription elongation by RNA polymerase II Supporting Evidence: PMID:22547686 The SEC family members demonstrate high levels of polymerase II (Pol II) C-terminal domain kinase activity PMID:31611376 we show that site-specific acetylation of super elongation complex (SEC) subunit AFF1 by p300 reduces its interaction with other SEC components and impairs P-TEFb-mediated C-terminal domain phosphorylation of RNA polymerase II both in vitro and in vivo |
| GO:0050877 nervous system process | IBA GO_REF:0000033 | KEEP AS NON CORE | Summary: Kept, but not as a core function. Both experimental donors reach this term from below: Drosophila lilli and mouse Aff2 each carry GO:0007611 learning or memory by IMP, a descendant, and PAINT generalised to the organ-system parent rather than transferring the specific term - which is the conservative and correct move, since neither learning nor memory has been assayed for human AFF1. There is no human AFF1 nervous-system experiment at all. Reason: AFF1's own nervous-system link is a mouse phenotype: the robotic allele carries a missense change in a conserved region of Af4 and the mice show region-specific Purkinje-cell loss with ataxia, with Igf-1 identified by ChIP as a target - attributed by those authors to "the AF4 transcriptional regulatory complex" rather than to Af4 alone. Stated at the strength the papers state it: the 2003 study is explicit that the causal step is hypothesised rather than demonstrated. That evidence is real but it is (i) mouse, (ii) a dominant missense allele in which Af4 accumulates, so a gain-of-function rather than a loss of function, and (iii) downstream of the same transcription-elongation activity captured by the core rows. The discriminator against core function 3, which uses similar tissue-specific-output language and is nevertheless kept as core, is the evidence and not the tissue-specificity: the DKK1 axis rests on direct human loss- AND gain-of-function in the relevant cell type with a mapped intermediate whose removal abolishes the effect, whereas this row has no human data at all and its mouse support is a dominant gain-of-function allele with no identified mechanism. MARK_AS_OVER_ANNOTATED is a defensible reading on the same facts and a curator may prefer it; KEEP_AS_NON_CORE is chosen because the term is not wrong - PAINT deliberately generalised from two independent learning-or-memory IMPs rather than transferring them, which is sound practice - and over-annotated would mis-describe a correct conservative call as an over-reach. Note for upstream curation: mouse Aff1 is NOT among this row's donors and MGI has annotated no nervous-system term to it, so the claim reaches human AFF1 through its paralogues while the ortholog's own relevant phenotype is uncaptured. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: FB:FBgn0041111 · Drosophila lilli (Q9VQI9) - the single fly family member, co-orthologue of all four vertebrate AFFs SUPPORTS TRANSFER the single fly family member, with its own experimental annotation in every subtree it is cited for MGI:MGI:1202294 · mouse Aff2 (O55112) - paralog SUPPORTS TRANSFER a paralog; its experimental evidence in this subtree is GO:0007611 learning or memory IMP (PMID:11923441), a descendant of the propagated term PANTHER:PTN000829417 · PANTHER family node (AF4/FMR2 family; 79 recipient gene products, human reach = AFF1/AFF2/AFF3/AFF4) SUPPORTS TRANSFER an internal PANTHER tree node, not a protein; its human reach is exactly the four AFF paralogues, all of which belong to SEC-family complexes Supporting Evidence: PMID:12629167 Genetic and physical mapping of the disease locus led to the identification of a missense mutation in a highly conserved region of Af4, a putative transcription factor that has been previously implicated in leukemogenesis. PMID:12629167 We demonstrate that Af4 is specifically expressed in Purkinje cells, and we hypothesize that the expression of mutant Af4 leads to neurodegeneration. PMID:20007461 Chromatin immunoprecipitation confirmed that Igf-1 is a direct and the first validated target of the AF4 transcriptional regulatory complex |
| GO:0032783 super elongation complex | IBA GO_REF:0000033 | ACCEPT | Summary: Accepted as a core cellular component. Unusually, the term's own definition names the family - a SEC contains "a transcription factor of the ELL family, an EAF protein, and an AFF family protein or distant relative" - so reading the definition supports this propagation rather than undermining it. The donor, Drosophila lilli, holds the term by IPI from PMID:22195968. All four human AFF paralogues receive it from this node, and that is right: AFF1 and AFF4 occupy SEC, AFF2 and AFF3 the biochemically isolated SEC-L2 and SEC-L3 variants. Reason: SEC membership is AFF1's central, best-corroborated fact: UniProt curates it from two references, it is reproduced by direct purification in human cells, and a point change in AFF1 alone is sufficient to disassemble the complex. The IBA and the independent human IDA agree, so this is corroboration rather than duplication. Propagation Review Root cause: NO FAILURE CORE Sources checked: FB:FBgn0041111 · Drosophila lilli (Q9VQI9) - the single fly family member, co-orthologue of all four vertebrate AFFs SUPPORTS TRANSFER the single fly family member, with its own experimental annotation in every subtree it is cited for PANTHER:PTN000829417 · PANTHER family node (AF4/FMR2 family; 79 recipient gene products, human reach = AFF1/AFF2/AFF3/AFF4) SUPPORTS TRANSFER an internal PANTHER tree node, not a protein; its human reach is exactly the four AFF paralogues, all of which belong to SEC-family complexes Supporting Evidence: file:human/AFF1/AFF1-uniprot.txt composed of EAF1, EAF2, CDK9, MLLT3/AF9, AFF (AFF1 or AFF4), the P-TEFb PMID:20159561 Furthermore, the ELL and AFF4-containing complexes also consist of additional MLL partners, AFF1, ENL, and AF9 PMID:31611376 Human SEC was described as a megadalton complex containing elongation factors P-TEFb (a heterodimer of CyclinT1 and CDK9) and ELL in association with AFF1 or AFF4, AF9 or ENL, and EAF1/2 |
| GO:0005634 nucleus | IEA GO_REF:0000120 | ACCEPT | Summary: Accepted. The location is correct and is independently established by the IDA row from PMID:41062835 and by nuclear immunofluorescence of both AF4 and its mouse ortholog. Worth recording about the provenance rather than the conclusion: GO_REF:0000120 asserts agreement between independent automatic pipelines, and one of its three tokens - UniProtKB-SubCell:SL-0191 - derives from this entry's own SUBCELLULAR LOCATION line, which is itself ECO:0000305 curator inference, so the three witnesses are not fully independent. Reason: A nuclear transcription-elongation scaffold has to be nuclear, and the experimental row for the same term makes this one redundant rather than load-bearing. Not flagged as a defect: the conclusion is right and the combinatorial reference is behaving as designed. Propagation Review Root cause: NO FAILURE CORE Sources checked: ARBA:ARBA00026330 · ARBA automatic rule SUPPORTS TRANSFER automatic rule, not a gene product InterPro:IPR043640 · InterPro AF4/FMR2 C-terminal homology domain signature SUPPORTS TRANSFER family-specific signature, not a gene product UniProtKB-SubCell:SL-0191 · UniProt SubCell keyword Nucleus (derived from the entry's own ECO:0000305 SUBCELLULAR LOCATION line) CIRCULAR OR REDUNDANT this token derives from the entry's own SUBCELLULAR LOCATION line, which is itself ECO:0000305 curator inference, so it is not an independent witness Supporting Evidence: file:human/AFF1/AFF1-uniprot.txt CC -!- SUBCELLULAR LOCATION: Nucleus {ECO:0000305}. |
| GO:0010468 regulation of gene expression | IEA GO_REF:0000002 | MODIFY | Summary: Sound but one level too coarse. This comes from InterPro2GO via IPR007797, the AF4/FMR2 family signature. PAINT gives that same family GO:0006355 - one level more specific - so two automatic pipelines assign different granularities to the same family from the same evidence base, and GO:0010468 was verified to be a genuine ancestor of GO:0006355, making it strictly redundant on this gene. Reason: The argument is upstream, about the mapping, not about this gene. Every characterised member of the AF4/FMR2 family regulates transcription specifically, not gene expression at some unspecified step - none has a reported role in translation, RNA stability or any other post-transcriptional layer, and the family's defining biochemistry is occupancy of a transcription-elongation complex. So IPR007797 supports the more precise term for every protein it matches, and the mapping is what should move. On THIS gene the row is redundant either way - today it is an ancestor of the GO:0006355 the gene already holds by IBA, and after the refinement it would duplicate that row exactly - so no gene-level redundancy is removed by the change, and the family-wide precision gain is the whole of the case for it. Propagation Review Root cause: NO FAILURE CORE Failure modes: GRANULARITY MISMATCH Sources checked: InterPro:IPR007797 · InterPro AF4/FMR2 family signature SUPPORTS TRANSFER family-specific signature, not a gene product Proposed replacements: regulation of DNA-templated transcription Supporting Evidence: PMID:22547686 We previously purified the AFF1- and AFF4-containing super elongation complex (SEC) as a major regulator of development and cancer pathogenesis. |
| GO:0005515 protein binding | IPI PMID:21729782 Human mediator subunit MED26 functions as a docking site for... | MODIFY | Summary: Replace with an informative molecular function. The partner is MLLT3/AF9 (P42568), captured here in the MED26 study's co-purification of SEC with Mediator; the reference-projection test returns 26 annotations over 17 entities, all GO:0005515, so this is one interaction network rather than a targeted binary assay. Judged per partner rather than per gene: this row and the PMID:23260655 row name the same partner and get the same verdict, while the HIV-1 Tat row is decided separately. Reason: AF9 is a verified partner by every available standard: five orthogonal detection methods in IntAct (NMR, circular dichroism, gel filtration, pull-down, tagged co-immunoprecipitation) across two publications, a deposited NMR structure (PDB 2LM0), and a dissociation constant of 0.17 nM - an order of magnitude tighter than AF9's other known ligands measured in the same experiment. So the problem with the row is not the partner, it is that bare protein binding records nothing about what the interaction accomplishes. What it accomplishes is bridging: AFF1 contacts AF9 through residues 761-774 and the P-TEFb kinase module through a separate region, and a single acetyl-mimic substitution in AFF1 is sufficient to lose SEC assembly and Pol II CTD phosphorylation. That is adaptor activity as GO defines it - bringing two or more macromolecules into contact so they can act in a coordinated way. This row contributes the in-cell half of that case: AFF1 and AF9 co-purify with Mediator on chromatin, which is where the bridging matters. Recorded for a curator's eye rather than as an objection: the same reference set led UniProt to assign GO:0060090 molecular adaptor activity to AF9 itself. That is not double counting - AF9's AHD is the hub competed for by Dot1L, BCoR and hPC3, while AFF1 is the hub joining P-TEFb to the ELL/EAF module - but the symmetry is worth checking. Proposed replacements: protein-macromolecule adaptor activity Supporting Evidence: PMID:21729782 Reactivation of paused Pol II correlates with recruitment of super-elongation complexes (SECs) containing ELL/EAF family members, P-TEFb, and other proteins PMID:31611376 Reexpression of wild-type AFF1, but not an acetylation mimic mutant, restores SEC component recruitment and target gene expression in AFF1 knockdown cells. |
| GO:0005515 protein binding | IPI PMID:22190034 Global landscape of HIV-human protein complexes. | KEEP AS NON CORE | Summary: Kept as a real but non-core host-pathogen interaction. The partner is HIV-1 Tat (P04608), from a systematic affinity-purification survey of all 18 HIV-1 proteins; the projection test returns 183 annotations over 104 distinct entities, all GO:0005515. Expanding the IntAct records rather than trusting UniProt's NbExp=3: the three records come from ONE publication, logged as two sub-methods in two cell lines (HEK293T and Jurkat) at MI-score 0.56. So NbExp=3 is again counting sub-methods, not independent experiments. Reason: The interaction is not screen noise: AFF1's role in Tat-dependent HIV transactivation is independently established, and the two-cell-line replication is part of that study's design. But it is not a core cellular function of AFF1, and the molecular detail is genuinely open - the corroborating study's stated mechanism is that AFF1 raises Tat's affinity for cyclin T1, which does not establish a direct AFF1-Tat contact, and AFF1 has no Tat co-structure (2LM0, its only PDB entry, is the AF9 complex). Bare protein binding is uninformative, but no better molecular-function term exists for a virus-host bridging role, so this is kept rather than modified. Supporting Evidence: PMID:22190034 we report the use of affinity tagging and purification mass spectrometry to determine systematically the physical interactions of all 18 HIV-1 proteins and polyproteins with host proteins in two different human cell lines (HEK293 and Jurkat) PMID:24367103 By increasing the affinity of the HIV-encoded transactivating (Tat) protein for CycT1, AFF1 facilitates Tat's extraction of P-TEFb from 7SK snRNP and the formation of Tat-SECs for HIV transcription. |
| GO:0005515 protein binding | IPI PMID:23260655 Leukemia fusion target AF9 is an intrinsically disordered tr... | MODIFY | Summary: Replace with an informative molecular function. Same partner as the PMID:21729782 row, MLLT3/AF9 (P42568), and this is the reference that quantifies it. The paper is titled for AF9, not AFF1 - a reminder that the only direct molecular measurement on human AFF1 sits under a partner's title. What was assayed is stated plainly: a 14-residue AFF1 peptide, not the full-length protein. For a coupled folding-and-binding motif that is the biologically meaningful unit rather than a truncation artefact, since AFF1's residues are disordered until they fold into AF9's hydrophobic core. Reason: AF9 is a verified partner by every available standard: five orthogonal detection methods in IntAct (NMR, circular dichroism, gel filtration, pull-down, tagged co-immunoprecipitation) across two publications, a deposited NMR structure (PDB 2LM0), and a dissociation constant of 0.17 nM - an order of magnitude tighter than AF9's other known ligands measured in the same experiment. So the problem with the row is not the partner, it is that bare protein binding records nothing about what the interaction accomplishes. What it accomplishes is bridging: AFF1 contacts AF9 through residues 761-774 and the P-TEFb kinase module through a separate region, and a single acetyl-mimic substitution in AFF1 is sufficient to lose SEC assembly and Pol II CTD phosphorylation. That is adaptor activity as GO defines it - bringing two or more macromolecules into contact so they can act in a coordinated way. This row contributes the quantitative and structural half: sub-nanomolar affinity by fluorescence anisotropy against MBP-AF9 AHD, a solved complex, and heteronuclear NOE data showing that only AFF1 residues 761-775 order on binding. Proposed replacements: protein-macromolecule adaptor activity Supporting Evidence: PMID:23260655 The affinity for AF4 is extremely high (KD = 0.17 ± 0.05 nM) PMID:23260655 we titrated a peptide derived from the AF9 interaction motif of AF4 (residues 761 to 774) into the AF9 AHD PMID:23260655 1H-15N heteronuclear NOE experiments show that AF4 residues 761-775 are ordered in the complex whereas the remainder is flexible PMID:23260655 aliphatic residues, which are conserved in each of the AF9 binding partners, form an integral part of the hydrophobic core of the complex file:human/AFF1/AFF1-uniprot.txt RP STRUCTURE BY NMR OF 738-779 IN COMPLEX WITH MLLT3. |
| GO:0000785 chromatin | IDA PMID:41062835 Stabilization of AFF1 by PARylation ensures transcriptional ... | ACCEPT | Summary: Accepted. This reference's cached record is abstract-only, so the direct assay is not visible here and the review defers to the curator, who read the full text. The call is corroborated from two independent directions: AFF1 occupies the DKK1 promoter by ChIP in mesenchymal stromal cells, and IntAct holds a ChIP record placing AFF1 at the MYC gene. Reason: Chromatin association follows from what AFF1 does - SEC acts on paused polymerase at promoters - and it is separately measured. Verified rather than assumed: GO:0000785 is not an is_a/part_of descendant of GO:0005634, so the chromatin and nucleus rows are two independent location claims and neither makes the other redundant. Supporting Evidence: PMID:28955517 we performed an anti-AFF1 ChIP assay, which demonstrated that AFF1 bound to the promoter region of DKK1 PMID:41062835 Upon DNA damage, PARP1 binds to and PARylates AFF1 in a region targeted by the E3 ligase Siah1, preventing AFF1 ubiquitination and promoting its stability. This stabilization supports efficient transcriptional recovery after DNA damage. |
| GO:0003711 transcription elongation factor activity | IMP PMID:41062835 Stabilization of AFF1 by PARylation ensures transcriptional ... | ACCEPT | Summary: Accepted, and it is already the most specific term available - GO:0003711 has no children, so there is no polymerase-resolved elongation-factor-activity term to refine to. Recorded because GO's evidence codes do not distinguish it: the evidence is a knockdown, so this establishes that AFF1 is REQUIRED for the activity, not that AFF1 alone is sufficient to supply it. Reason: The stimulation of polymerase elongation is delivered by the assembled complex - the CTD kinase is CDK9, not AFF1 - so core_functions records this under contributes_to_molecular_function rather than as an activity AFF1 carries out alone. The independent basis for AFF1's specific contribution is the acetyl-mimic experiment: re-expressing wild-type but not acetylation-mimic AFF1 restores SEC recruitment and target-gene expression, so a change in AFF1 alone controls whether the elongation activity is delivered. Verified rather than assumed: GO:0003712 is NOT an ancestor of GO:0003711, so this row and the coregulator row are different claims rather than a general/specific pair, and both stand. The divergence between the row's GOA qualifier (enables) and where core_functions puts the term (contributes_to_molecular_function) is deliberate, not an inconsistency: strictly, contributes_to is the correct GOA qualifier for an activity delivered by an assembled complex, but ActionEnum offers no qualifier-change action, and MODIFY would wrongly imply the TERM is wrong when only the qualifier is imprecise. So the row is ACCEPTed as-is and the nuance is carried by the core_functions slot; a qualifier correction is raised with UniProt in suggested_questions instead. Supporting Evidence: PMID:41062835 AFF1 depletion impairs DNA repair and survival PMID:31611376 Reexpression of wild-type AFF1, but not an acetylation mimic mutant, restores SEC component recruitment and target gene expression in AFF1 knockdown cells. PMID:22547686 The SEC family members demonstrate high levels of polymerase II (Pol II) C-terminal domain kinase activity |
| GO:0005634 nucleus | IDA PMID:41062835 Stabilization of AFF1 by PARylation ensures transcriptional ... | ACCEPT | Summary: Accepted. This is the experimental nuclear localisation for AFF1; note that UniProt's own SUBCELLULAR LOCATION line is ECO:0000305 curator inference, so before this reference the nuclear call rested on inference plus older immunofluorescence rather than on a curated experimental annotation. Reason: Uncontested and corroborated by the IEA row for the same term and by nuclear immunofluorescence of AF4 and of mouse Af4. Supporting Evidence: file:human/AFF1/AFF1-uniprot.txt CC -!- SUBCELLULAR LOCATION: Nucleus {ECO:0000305}. PMID:41062835 Upon DNA damage, PARP1 binds to and PARylates AFF1 in a region targeted by the E3 ligase Siah1, preventing AFF1 ubiquitination and promoting its stability. This stabilization supports efficient transcriptional recovery after DNA damage. |
| GO:0006974 DNA damage response | IDA PMID:41062835 Stabilization of AFF1 by PARylation ensures transcriptional ... | ACCEPT | Summary: Accepted. The projection test on this reference was run precisely because AFF1 is a complex subunit and a complex-level phenotype can become gene-level evidence on every subunit: fully paginated, PMID:41062835 yields 8 annotations over 2 distinct gene products, with six terms on AFF1 alone and only the damage-site location shared with PARP1. The phenotype does not spread across a complex, so this is the benign shape. Reason: Independently corroborated from the opposite direction by the p300 study, which shows AFF1 acetylation is dynamically induced by genotoxic stress and shuts SEC down. Two laboratories, two modifications, one coherent picture: AFF1 abundance and modification state are the switch for transcriptional shutdown and restart around DNA damage. Supporting Evidence: PMID:41062835 Upon DNA damage, PARP1 binds to and PARylates AFF1 in a region targeted by the E3 ligase Siah1, preventing AFF1 ubiquitination and promoting its stability. This stabilization supports efficient transcriptional recovery after DNA damage. PMID:41062835 AFF1 depletion impairs DNA repair and survival PMID:31611376 we show that site-specific acetylation of super elongation complex (SEC) subunit AFF1 by p300 reduces its interaction with other SEC components and impairs P-TEFb-mediated C-terminal domain phosphorylation of RNA polymerase II both in vitro and in vivo |
| GO:0032783 super elongation complex | IDA PMID:20159561 AFF4, a component of the ELL/P-TEFb elongation complex and a... | ACCEPT | Summary: Accepted as a core cellular component, after checking what molecule the experiment used. This paper is titled for AFF4 and most of it works with MLL-AFF1 and other MLL-fusion constructs, which would be the wrong basis for a wild-type annotation. This row is not that: its basis is the Flag-AFF4 and Flag-ELL1/2/3 purifications, which recover endogenous AFF1 as a component. Wild-type protein, native levels. Reason: SEC membership is AFF1's core cellular component and this is the direct human biochemical demonstration of it. The reference-projection test returns a single annotation on a single entity, so there is no projection concern. The same paper independently records that AFF1 and AFF4 are not interchangeable - depleting AFF1 does not destabilise ELL1 or P-TEFb, whereas depleting AFF4 does - which is why this review treats the paralogues separately throughout. Supporting Evidence: PMID:20159561 Furthermore, the ELL and AFF4-containing complexes also consist of additional MLL partners, AFF1, ENL, and AF9 PMID:20159561 We observed that the reduction of the AFF4 homologue AFF1 does not alter ELL1 and P-TEFb stability in these cells file:human/AFF1/AFF1-uniprot.txt composed of EAF1, EAF2, CDK9, MLLT3/AF9, AFF (AFF1 or AFF4), the P-TEFb |
| GO:0032786 positive regulation of DNA-templated transcription, elongation | IMP PMID:41062835 Stabilization of AFF1 by PARylation ensures transcriptional ... | MODIFY | Summary: Correct but redundant with its own child. GO:0032968 was verified by fetching its is_a/part_of closure to be a descendant of GO:0032786, and the SAME reference supplies both by IMP - so these two rows are a parent/child pair from one experiment, not two independent findings. AFF1's substrate is RNA polymerase II, so the polymerase-agnostic parent carries no information the child does not. Reason: Collapse onto the Pol II-resolved child that the same reference already provides. Every measurement behind these rows is Pol II-specific: SEC-family complexes are assayed as Pol II CTD kinases and the acetylation study measures Pol II CTD phosphorylation. There is no evidence for AFF1 acting on RNA polymerase I or III, so nothing is lost. Deliberately NOT claimed: that GO:0032968 sits under GO:0006368; it does not - GO links them by positively_regulates and keeps regulation out of the is_a hierarchy, and both closures were fetched to check. Proposed replacements: positive regulation of transcription elongation by RNA polymerase II Supporting Evidence: PMID:41062835 AFF1 depletion impairs DNA repair and survival PMID:22547686 The SEC family members demonstrate high levels of polymerase II (Pol II) C-terminal domain kinase activity |
| GO:0032968 positive regulation of transcription elongation by RNA polymerase II | IMP PMID:41062835 Stabilization of AFF1 by PARylation ensures transcriptional ... | ACCEPT | Summary: Accepted as a core biological process, and it is the term the parent row should collapse onto. AFF1 knockdown lowers transcriptional recovery after DNA damage, and AFF1's contribution to Pol II CTD phosphorylation is separately demonstrated by the acetyl-mimic experiment. Reason: This is the polymerase-resolved statement of AFF1's central activity and it is supported from both directions - loss of AFF1 lowers elongation output, and restoring wild-type but not acetylation-mimic AFF1 recovers it. The evidence is knockdown-based, so it establishes requirement; the sufficiency arm comes from the rescue rather than from this row. Supporting Evidence: PMID:41062835 AFF1 depletion impairs DNA repair and survival PMID:31611376 we show that site-specific acetylation of super elongation complex (SEC) subunit AFF1 by p300 reduces its interaction with other SEC components and impairs P-TEFb-mediated C-terminal domain phosphorylation of RNA polymerase II both in vitro and in vivo PMID:31611376 Reexpression of wild-type AFF1, but not an acetylation mimic mutant, restores SEC component recruitment and target gene expression in AFF1 knockdown cells. |
| GO:0090734 site of DNA damage | IDA PMID:41062835 Stabilization of AFF1 by PARylation ensures transcriptional ... | ACCEPT | Summary: Accepted, deferring to the curator on an assay this review cannot see - the cached record for this reference is abstract-only. The abstract does establish the biochemical basis for a damage-site location: PARP1, which localises to DNA lesions, binds and PARylates AFF1. The projection test shows this is the only term of the seven that PMID:41062835 gives to a second entity, and that entity is PARP1 itself - consistent with a shared recruitment rather than a complex-wide projection. Reason: A damage-induced location for a transcription-elongation scaffold is exactly what the transcription-restart mechanism requires, and it is recorded in core_functions as a location of the DNA-damage-response function rather than of the constitutive SEC function, since it is stress-induced. Verified rather than assumed: GO:0090734 is not an is_a/part_of descendant of GO:0000785, so this and the chromatin row are separate claims. Supporting Evidence: PMID:41062835 Upon DNA damage, PARP1 binds to and PARylates AFF1 in a region targeted by the E3 ligase Siah1, preventing AFF1 ubiquitination and promoting its stability. This stabilization supports efficient transcriptional recovery after DNA damage. |
| GO:0006354 DNA-templated transcription elongation | EXP PMID:22547686 The super elongation complex family of RNA polymerase II elo... | MODIFY | Summary: Same term and same granularity issue as the IBA row, and given the same action for consistency. This is the human experimental row for AFF1's participation in transcription elongation, from the study that biochemically isolated the SEC family; the reference-projection test returns one annotation on one entity, so there is no projection concern. Reason: The very reference behind this row is what makes the polymerase specific: it reports that the SEC family members are Pol II C-terminal domain kinases. GO:0006368 was verified to be an is_a/part_of descendant of GO:0006354 before being proposed, so this is a one-step refinement toward the polymerase the experiment actually used. Proposed replacements: transcription elongation by RNA polymerase II Supporting Evidence: PMID:22547686 We previously purified the AFF1- and AFF4-containing super elongation complex (SEC) as a major regulator of development and cancer pathogenesis. PMID:22547686 The SEC family members demonstrate high levels of polymerase II (Pol II) C-terminal domain kinase activity |
| GO:0008023 transcription elongation factor complex | IDA PMID:22195968 The little elongation complex regulates small nuclear RNA tr... | MODIFY | Summary: Correct but one level less specific than the same curator's own reading of the same paper. UniProt's reference table for PMID:22195968 records its contribution as identification in the SEC complex, yet the annotation landed on the generic parent. The reference-projection test makes the pattern exact. Twenty-two gene products receive one of the two terms, 11 human and 11 Drosophila. All 10 recipients of the specific GO:0032783 are Drosophila, and ALL ELEVEN human recipients - AFF1, AFF4, CDK9, ELL, ELL2, ELL3, EAF1, EAF2, MLLT1, MLLT3 and ICE2 - got only the generic GO:0008023. The split is not purely clade-based: Drosophila Ell is the twelfth parent-only recipient, so 10 of the 11 fly subunits were refined and none of the 11 human ones was. Reason: GO:0032783 was verified to be an is_a/part_of descendant of GO:0008023, and the specific term is already established for AFF1 by an independent human IDA, so this refinement asserts nothing new about the gene. What makes it worth acting on is the route the specific term took to get here: Drosophila lilli's GO:0032783 IPI comes from THIS paper, and that fly annotation is the donor of human AFF1's GO:0032783 IBA. So the specific term reached AFF1 by phylogenetic inference from a fly protein annotated in the same experiment that annotated the human protein one level up. Fixing it at the reference would retract the detour for all eleven human recipients at once; that recommendation is stated once, in suggested_questions, rather than repeated per gene. Proposed replacements: super elongation complex Supporting Evidence: file:human/AFF1/AFF1-uniprot.txt RP IDENTIFICATION IN THE SEC COMPLEX. PMID:22195968 Eleven-nineteen lysine-rich leukemia (ELL) participates in the super elongation complex (SEC) with the RNA polymerase II (Pol II) CTD kinase P-TEFb. PMID:20159561 Furthermore, the ELL and AFF4-containing complexes also consist of additional MLL partners, AFF1, ENL, and AF9 |
| GO:0045668 negative regulation of osteoblast differentiation | IMP PMID:28955517 AFF1 and AFF4 differentially regulate the osteogenic differe... | NEW | Summary: Proposed as a new annotation. AFF1's best-characterised gene-specific human function has no representation in GOA at all. In human mesenchymal stromal cells, siRNA depletion of AFF1 increases alkaline phosphatase staining, mineralisation and Runx2/SP7/BGLAP expression, while lentiviral overexpression decreases them and reduces ectopic bone formation in vivo. AFF1 occupies the DKK1 promoter, sets DKK1 levels in both directions, and knocking DKK1 down abolishes the effect of AFF1 overexpression - so the intermediate is identified, not merely correlated. Reason: Both directions were tested, which matters because GO's evidence codes do not distinguish them: the knockdown arm shows AFF1 is REQUIRED to restrain osteogenesis and the overexpression arm shows raising AFF1 is SUFFICIENT to suppress it. IMP is the right code for both, since the assays are siRNA depletion and ectopic expression rather than direct measurement at native levels. Sign matters here for a family reason: the same paper shows AFF4 has the OPPOSITE effect, so a term transferred between the two paralogues in either direction would invert the biology. Two caveats recorded rather than smoothed over - the paper carries a 2020 Correction (PMID:32257529) whose content is not stated in any retrievable record, and one sentence in its overexpression section says "AFF1-depleted" where the surrounding figure describes overexpressing cells; the claims above rest only on the unambiguous sentences. Supporting Evidence: PMID:28955517 siRNA-mediated depletion of AFF1 led to more intense staining of alkaline phosphatase (ALP), an early marker of osteoblastic differentiation PMID:28955517 We found that overexpression of AFF1 decreased the ALP activity and mineralization of MSCs PMID:28955517 we performed an anti-AFF1 ChIP assay, which demonstrated that AFF1 bound to the promoter region of DKK1 PMID:28955517 Depletion of DKK1 significantly abolished the inhibition of ALP activity triggered by AFF1 overexpression PMID:28955517 Mice implanted with AFF1-overexpressing MSCs showed much less bone tissue PMID:28955517 depletion of AFF4 significantly reduced the alkaline phosphatase (ALP) activity and extracellular matrix mineralization, indicating that it had an opposite effect on osteogenic differentiation compared with AFF1 |
Loading supporting content…
Download this section (compressed HTML)Q: PAINT/UniProt curation, stated once for all affected genes: PMID:22195968 gives GO:0032783 super elongation complex to 10 Drosophila recipients and only the generic parent GO:0008023 to all eleven of its human recipients - AFF1, AFF4, CDK9, ELL, ELL2, ELL3, EAF1, EAF2, MLLT1, MLLT3 and ICE2 - even though UniProt's own reference table records the paper's contribution as "IDENTIFICATION IN THE SEC COMPLEX". The consequence is a detour: Drosophila lilli's GO:0032783 IPI from this paper is the donor of human AFF1's GO:0032783 IBA, so the specific term reaches the human gene by phylogenetic inference from a fly protein annotated in the same experiment. Would GOA consider refining the human rows at the reference, which would fix eleven gene products in one edit? Note the split is not purely clade-based - Drosophila Ell also received only the parent - so this is a per-annotation gap rather than a rule about clades.
Suggested experts: GOA/UniProt curators, GO Central PAINT curators
Q: PANTHER node PTN000829417 assigns, to the same 79 recipients, both GO:0032783 - whose definition is explicitly about RNA polymerase II elongation - and the polymerase-agnostic GO:0006354. Since every characterised SEC substrate is Pol II, would PAINT consider moving the process term to GO:0006368 transcription elongation by RNA polymerase II at the node, which would refine all four human AFF paralogues at once? Separately, InterPro2GO gives the same family GO:0010468 via IPR007797 while PAINT gives GO:0006355, so the two pipelines differ by one level on the same family.
Suggested experts: GO Central PAINT curators, InterPro2GO curators
Q: MGI curation gap: the AFF1-specific nervous-system literature is a mouse phenotype - the robotic allele stabilises Af4 and causes Purkinje-cell loss with ataxia (PMID:12629167), with Af4 regulating Igf-1 in Purkinje cells (PMID:20007461) - yet mouse Aff1 carries no nervous-system term, and human AFF1's GO:0050877 IBA therefore rests on the paralogues Aff2 and lilli instead of on the ortholog. Would MGI consider annotating the robotic phenotype?
Suggested experts: MGI curators
Q: Do AFF1 and AFF4 ever occupy the same super elongation complex? PMID:28955517's discussion states they "are not found together in a single SEC" and UniProt models them as alternative occupants of one slot ("AFF (AFF1 or AFF4)"), but IntAct records an AFF1-AFF4 physical association across 8 records and 2 distinct publications at MI-score 0.6, and PMID:20159561's Flag-AFF4 purification recovers endogenous AFF1. If they are genuinely mutually exclusive, GO cannot currently say so - GO:0032783 has no children and ComplexPortal's seven AFF1-containing entries all list AFF1 and AFF4 together.
Suggested experts: transcription elongation biologists, ComplexPortal curators
Q: UniProt: would a FUNCTION comment be considered for P51825? The entry currently has none, despite a solved partner complex, a measured dissociation constant, a curated SUBUNIT line and a 2026 primary paper on its role in transcriptional recovery. The SUBCELLULAR LOCATION line is also still ECO:0000305 curator inference although an experimental IDA now exists (PMID:41062835). Third item, raised here because ActionEnum has no qualifier-change action so it cannot be expressed on the row itself: GO:0003711 transcription elongation factor activity is annotated with the enables qualifier, but the activity is delivered by the assembled SEC and the evidence is an AFF1 knockdown, so contributes_to would be the strictly correct qualifier. The term is right; only the qualifier overstates AFF1's independence.
Suggested experts: UniProt curators
Q: Is AFF1's molecular adaptor role better modelled on AFF1, on AF9, or on both? UniProt assigned GO:0060090 molecular adaptor activity to AF9 (P42568) from PMID:23260655, and this review proposes GO:0030674 for AFF1 from the same interaction plus the acetyl-mimic rescue. Both proteins are hubs, but for different ligand sets, and a curator's view on whether adaptor activity should be annotated reciprocally would be useful.
Suggested experts: GO Central curators, UniProt curators
Experiment: Sequential immunodepletion followed by quantitative mass spectrometry: deplete AFF1 from nuclear extract to exhaustion, then ask whether the remaining ELL/EAF/P-TEFb/AF9 pool still contains AFF4 at near-stoichiometric levels, and repeat in the opposite order. Pair it with single-molecule two-colour photobleaching on complexes purified through a third subunit (ELL2 or EAF1) to count AFF1 and AFF4 copies per particle. This distinguishes genuine co-occupancy from an average over two populations, which is what a bulk co-immunoprecipitation cannot do and which is why the current evidence is contradictory.
Hypothesis: AFF1-containing and AFF4-containing super elongation complexes are mutually exclusive assemblies rather than one complex carrying both.
Type: biochemistry
Experiment: Knock in, at the endogenous locus, the acetylation-mimic and acetylation-defective AFF1 alleles used in PMID:31611376 together with a PARylation-site mutant of the Siah1-targeted region from PMID:41062835, then measure recovery of nascent transcription after UV or camptothecin by TT-seq alongside AFF1 protein levels and SEC composition. The informative outcome is a separation: an allele that keeps normal AFF1 abundance but fails to recover transcription would show the defect is assembly, not dosage.
Hypothesis: The bridging function of AFF1 is separable from its abundance, so the transcriptional-restart phenotype after DNA damage is caused by loss of SEC assembly rather than by loss of AFF1 protein per se.
Type: genetics
Experiment: In one human mesenchymal stromal cell background, perform paired degron-mediated acute depletion of AFF1 and of AFF4 with TT-seq and Pol II PRO-seq readouts, plus ChIP-seq for each paralogue. Then swap the AF9-binding motifs between them and ask whether target selection follows the motif or the rest of the protein. A negative control is built in: if the two depletions converge on the same genes, the differential osteogenesis result requires another explanation.
Hypothesis: AFF1 and AFF4 direct SEC to non-overlapping gene sets, and the osteogenic sign inversion between them follows from that target split rather than from different biochemistry.
Type: genomics
Experiment: Measure binding of recombinant Tat to purified AFF1 fragments by isothermal titration calorimetry and NMR chemical-shift perturbation, with and without cyclin T1 present, and attempt a co-structure of the AFF1 region that is not the AF9-binding motif. AFF4-Tat-P-TEFb structures exist and AFF1 has none, so the honest current position is that the GO:0005515 row rests on affinity purification and a functional requirement, not on a mapped contact.
Hypothesis: AFF1 contacts HIV-1 Tat directly, rather than only raising Tat's affinity for cyclin T1 within a ternary complex.
Type: structural_biology
Experiment: The original result used a GAL4-DNA-binding-domain chimera, so it shows a region can activate when artificially tethered and cannot show what AFF1 does natively - which is why no GO term is proposed from it here. Test it properly by tethering the same region with a catalytically dead dCas9 to endogenous AFF1 target promoters identified by ChIP-seq, and by deleting that region at the endogenous locus and measuring nascent transcription of those same promoters.
Hypothesis: The transactivation activity reported for an AF-4 GAL4 fusion in 1996 reflects a function the native protein performs at its own target promoters.
Type: genetics
What is not known — curated, literature-grounded statements of the open unknowns (the inverse of core functions).
Gap: Whether AFF1 and AFF4 can occupy the same super elongation complex is unresolved, and the evidence points both ways. UniProt models them as alternative occupants of one slot and PMID:28955517's discussion states they are not found together, but IntAct records an AFF1-AFF4 association across 8 records and 2 distinct publications at MI-score 0.6 and PMID:20159561's AFF4 purification recovers endogenous AFF1.
OPEN BIOLOGY CC_DARK
What is known: What is settled is that the two paralogues are not functionally interchangeable: depleting AFF1 does not destabilise ELL1 or P-TEFb whereas depleting AFF4 does, their SEC target genes are reported as largely non-overlapping, and in mesenchymal stromal cells they act in opposite directions on osteogenesis. What is not settled is complex stoichiometry, which bulk co-immunoprecipitation cannot decide because it averages over populations.
Significance: It determines whether GO:0032783 needs paralogue-specific children. No such children are proposed here, because proposing terms for a distinction the evidence does not settle would be an over-annotation of the opposite sign.
Provenance (the field's own admissions):
Gap: AFF1's subnuclear distribution is described inconsistently. Older immunofluorescence reports discrete punctate compartments ('AF4 bodies') where AF4 and AF9 co-localise, while PMID:28955517's discussion describes AFF1 as diffuse and reserves nuclear speckles for AFF4. GOA reflects the uncertainty by giving AFF1 no subnuclear term at all, while AFF4 has nucleoplasm and nuclear body and AFF2 has nuclear speck.
OPEN BIOLOGY CC_DARK
What is known: Nuclear localisation itself is settled, by IDA and by immunofluorescence in several cell types. Chromatin association is settled, by ChIP at DKK1 and at MYC. What is unresolved is whether AFF1 concentrates in a defined subnuclear body.
Significance: No subnuclear term is proposed for AFF1 here. The absence of one in GOA appears to be appropriate caution rather than a coverage gap.
Provenance (the field's own admissions):
Gap: Whether AFF1 has any molecular function outside a super elongation complex is unknown. Every measurement on the protein is made either within SEC or on a short peptide from it, and its only solved structure is 42 residues bound to a partner; 901 of its 1210 residues (74.5%) are annotated as disordered, most of it with no assigned interaction.
OPEN BIOLOGY MF_DARK
What is known: The SEC-dependent activity is well established and quantified. What is unmapped is the function of the large disordered regions outside the AF9-binding motif and the Siah1/PARylation region - including the nine MS-detected phosphosites and one acetylation site, none of which has an assigned consequence.
Significance: This is why GO:0030674 is proposed as the molecular function rather than anything narrower, and why the elongation-factor and coregulator terms are recorded as contributions to a complex.
Provenance (the field's own admissions):
Gap: What the wild-type AFF1 protein contributes to t(4;11) leukaemogenesis, as distinct from the MLL-AFF1 chimera, is not established. The chimera is a different molecule with its own composition, and the peptide inhibitor PFWT that kills t(4;11) cells is a synthetic reagent modelled on AFF1's AF9-binding domain rather than the gene product.
OPEN BIOLOGY BP_DARK
What is known: The AF9-binding motif itself is mapped and quantified on wild-type AFF1. What is not separable from the current data is which leukaemia phenotypes require the remaining wild-type allele's own activity.
Significance: No GOA row for AFF1 rests on the fusion or on the peptide, which is the correct state; this gap records why no leukaemia process term is proposed here.
Provenance (the field's own admissions):
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)