Oncofusion

start / graph / gene-nras

Genestrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q46-cooperating-lesions

NRAS - the other RAS, the one whose 0.3% clone came back, and a UniProt record that lists HRAS1 as its own synonym

also called NRAS · N-Ras · Transforming protein N-Ras · P01111

NRAS is the sister switch to KRAS and is mutated in the same small fraction of infant KMT2A leukaemias. The most important thing known about it here is not how often it is mutated but where the mutation hides: in one study, NRAS-mutant cells making up as little as three cells in a thousand at diagnosis were the ones that grew back at relapse.

The same thing, for a scientist

NRAS (P01111, 189 aa) is, with KRAS, one of the only two genes recurrently point-mutated in infant KMT2A-rearranged B-ALL [E2]. Its frequency is lower than KRAS in the whole-genome discovery cohort (2 of 22 cases against 4) [E1] and the two are usually discussed together, which is precisely why the graph had a pathway entry and no gene entry. The reason NRAS deserves a separate entry is a measurement that applies to it specifically. Targeted **duplex sequencing** of infant and childhood KMT2A-r ALL found drug-resistant clones carrying **NRAS G13D or G12S** - and FLT3 D835H - present at **0.06% to 0.34% of diagnostic cells**, which then expanded at relapse [E3]. That is two to three orders of magnitude below what the cohort studies that established the 'subclonal and lost at relapse' picture could detect. It does not overturn the low-mutation-burden claim; it changes what 'lost at relapse' is allowed to mean. NRAS behaves differently from KRAS in one published respect worth recording: in paediatric B-ALL, NRAS mutation correlates with KRAS mutation (phi 0.19) and with PTPN11 mutation (phi 0.138), i.e. RAS-pathway lesions co-occur in the same patients rather than being mutually exclusive [E5] - which for a set of lesions in the same linear pathway is the opposite of what a one-hit model predicts and is consistent with several of them sitting in separate subclones.

What it looks like
loading AF-P01111-F1…
AlphaFold AF-P01111-F1 ↗
colourshape

This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.

very high (90+)confident (70–90)low (50–70)very low (<50)
I have never read one of these pictures before

A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.

Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.

Anything in yellow sticks is a bound small molecule, which is usually the drug.

Go deeper
An identifier warning specific to this record

gene_exact:NRAS AND organism_id:9606 AND reviewed:true with gene_primary and length returns one reviewed human record - but read its synonym field:

accessionprimarygene synonymsprotein namelength
P01111NRASHRAS1GTPase NRas (Transforming protein N-Ras)189

UniProt lists HRAS1 as a gene synonym of NRAS. HRAS is a different RAS family member (P01112, 189 aa). This is the rule-63 shape - a curated alias pointing at a different gene in the same family - and here the paralogue trap is worse than usual on two counts. First, the three canonical RAS proteins are all 189 residues, so the length check that rescues STAG1-from-PMEPA1 (287 against 1,258) has no discriminating power at all here. Second, the codons are shared: a 'G12D' or 'G13D' is a legitimate annotation in all three genes, so a mis-assigned mutation reads as a perfectly ordinary RAS mutation.

The only defence is the primary gene name, and it is the only field this entry relies on. verify-accessions.py will catch an accession swap, and it will not catch NRAS-for-HRAS by length. Anyone extending this entry to HRAS must resolve the accession independently rather than by alias.

What duplex sequencing changed, and what it did not

The standing description of RAS lesions in this disease - subclonal at diagnosis, frequently lost at relapse 12 - was established with WGS at about 39x plus deep custom capture on five cases, and with Sanger sequencing. Duplex sequencing lowers the reliable detection floor to roughly 0.01-0.1% by requiring both strands of an original duplex molecule to agree.

At that floor, in 20 infants and 7 children with KMT2A-r ALL 3:

  • 62% of the identified cancer-associated mutations were below the resolution of standard NGS.
  • Only 33 of 87 exceeded 2% cellular prevalence, and 31 of those 33 were PI3K/RAS genes, typically KRAS or NRAS.
  • Five patients had only low-frequency PI3K/RAS mutations, with no higher-frequency signalling mutation at all - patients who would be scored RAS-wild-type by every cohort study cited above.
  • Drug-resistant clones with NRAS G13D or G12S (or FLT3 D835H) at 0.06-0.34% of diagnostic cells expanded at relapse.
  • In longitudinal samples the relapse clone persisted as a minor subclone from diagnosis, through treatment, and expanded only in the last month of disease.

What this does not do is overturn the quiet-genome claim. The same paper reports that infants had fewer such mutations than the older children 3. Depth moved the floor; it did not reverse the age effect. The right reading is that the burden figure is an instrument reading with a stated floor, and the subclonal reservoir beneath that floor is real, small, and prognostically relevant.

Frequency in this disease
cohortassayNRAS
22 infant MLL-R ALL, WGS 1WGS 39xrecurrent, n = 2 (KRAS n = 4)
124 infant B-ALL, Interfant-99/06 2genome-widerecurrent mutation exclusively in K-RAS and N-RAS
109 infant ALL 4Sanger, NRAS/KRAS/BRAFRAS about 14% overall, about 24% in t(4;11)
27 infant/child KMT2A-r ALL 3duplex + NGSNRAS among the recurrently altered genes; resistant NRAS clones at 0.06-0.34%
20 non-infant MLL-R leukaemias 1WGS/panelNRAS n = 3, and pathway mutations mostly in the major clone, unlike infants

The last row is the contrast that matters. In older children with the same fusion class, the same pathway mutations are clonal. Whatever makes these lesions subclonal is a property of the infant disease, not of the fusion.

Facts and figures
uniprotP01111
gene symbolNRAS
length AA189
alphafold IDAF-P01111-F1
cytoband1p13.2
resolution notegene_exact:NRAS AND organism_id:9606 AND reviewed:true with fields=accession,gene_primary,gene_synonym,protein_name,length returned one row: P01111, primary gene name NRAS, 189 aa, fetched from the UniProt REST API on 2026-09-20. WARNING recorded in the entry body: its only listed gene synonym is HRAS1, and HRAS (P01112) is also 189 aa, so length cannot discriminate NRAS from HRAS. Resolution here rests entirely on the primary gene name.
domains noteUniProt P01111 carries no Domain features; the G-domain and switch regions are annotated as regions and binding sites, so no domain is recorded rather than a guessed boundary.
druggabilitySame position as KRAS: no allele-selective NRAS inhibitor in clinical use in this disease, so the route is downstream MEK inhibition, with the same subclonal-target problem and the same age floor on the only open paediatric protocol.
subnuclear localizationNOT APPLICABLE - NRAS is a lipid-anchored membrane GTPase with no nuclear pool. Recorded as not-applicable, not as an unmeasured absence.
shown for fusionNO for localisation. Nobody has imaged NRAS, or measured its acylation state, in any KMT2A-rearranged cell. The bounded localisation query returned 2 records and neither is a localisation study: PMID 21148331 and PMID 15994285 are both AML prognosis papers (RUNX1 mutations in de novo AML; NPM1 mutations in AML) that mention NRAS and MLL in their abstracts. NRAS MUTATION in KMT2A-rearranged leukaemia is well documented and is recorded elsewhere in this entry; that is genotype, not localisation.
paralogue trap
{
  "kind": "rule-63 alias pointing at a family member, with the usual length defence unavailable",
  "detail": "P01111 (NRAS) lists HRAS1 as a gene synonym. KRAS P01116, NRAS P01111 and HRAS P01112 are all 189 aa, and hotspot codons 12, 13 and 61 are shared, so a mis-assignment produces an ordinary-looking RAS mutation in the wrong gene.",
  "defence": "match on gene_primary only; do not rely on length or on the hotspot codon"
}
domains
[]
cooperating lesion profile
{
  "disease": "infant KMT2A-rearranged B-ALL",
  "recurrence_wgs_discovery": "2 of 22 cases",
  "subclonal_reservoir": "NRAS G13D and G12S clones at 0.06-0.34% of diagnostic cells expanded at relapse",
  "co_occurrence": "correlated with KRAS (phi 0.19) and PTPN11 (phi 0.138) in paediatric B-ALL, i.e. not mutually exclusive",
  "age_contrast": "pathway mutations mostly in the MAJOR clone in non-infant MLL-R leukaemia (P < 0.0001 against infants)",
  "verdict": "subclonal marker whose clinical weight comes from the fraction of it that is clonal, plus a sub-0.5% reservoir that selects under therapy"
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.3075,
  "chronos_mean_other_leukaemia": -0.4414,
  "difference": 0.1339,
  "pan_essentiality_all_lines": -0.2207,
  "selectivity_rank": 17340,
  "rank_denominator": 17728,
  "provenance": "Joined 2026-09-21 from analyses/depmap-kmt2a-selective-scan-full (generated 2026-09-19T10:12:06Z), the project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies here, including that the KMT2A-rearranged arm is a small number of lines drawn from nine individuals, that the rank denominator is 17,728 genes, and that a difference is not a dependency. Added because only 26 of 308 gene entries carried this number, which made any KMT2A-selectivity question answerable for 8% of the corpus.",
  "how_to_read": "A negative difference means MORE required in KMT2A-rearranged lines than in other leukaemia lines. A rank near 17,728 is ANTI-selective. Neither is evidence of a fusion-driven dependency on its own: check pan_essentiality_all_lines before calling anything selective, and note that a paralogue-redundant gene can score near zero for reasons that have nothing to do with this disease - see ana-the-paralogue-redundancy-excuse-is-not-tested-and-my-proxy-decided-the-answer."
}
localization
{
  "subcellular": "PLASMA MEMBRANE and GOLGI APPARATUS MEMBRANE, both as a LIPID ANCHOR, shuttling between the two. UniProt P01111 records exactly this: Cell membrane (lipid-anchor) and Golgi apparatus membrane (lipid-anchor), both ECO:0000269 PubMed:15705808 and 26701913, with the note 'Shuttles between the plasma membrane and the Golgi apparatus'. HPA immunofluorescence: main Plasma membrane, Cytosol additional, reliability Supported.",
  "subnuclear": "NOT APPLICABLE. NRAS has no nuclear pool; it is a lipid-anchored membrane GTPase. Recorded as not-applicable rather than as an unmeasured absence, so that it is not counted as a missing measurement.",
  "THE_FIELD_THAT_MATTERS_IS_THE_ACYLATION_CYCLE": "A record that says 'cytoplasm' or even 'plasma membrane' misses the whole point for this gene. NRAS is farnesylated (a stable, irreversible modification) AND palmitoylated (a reversible one), and it is the palmitoylation/depalmitoylation cycle that decides the address: palmitoylation traps NRAS at the plasma membrane, depalmitoylation releases it back to the Golgi, and the cycle continuously maintains the plasma-membrane pool against equilibration (PMID 15705808). The ABHD17 family are the depalmitoylases that set N-Ras palmitate turnover and hence its subcellular localisation (PMID 26701913). Signalling output is a function of WHICH membrane NRAS is on, so localisation here is mechanism and not description.",
  "therapeutic_consequence": "This is why farnesyltransferase inhibitors failed against NRAS while the localisation biology remained real: NRAS can be geranylgeranylated when farnesylation is blocked, so blocking one prenylation route does not evict it from the membrane. The exploitable step, if there is one, is the reversible palmitoylation cycle rather than the stable prenylation - and zDHHC-directed and ABHD17-directed chemistry is an active area. Nothing of this has been tested in a KMT2A-rearranged cell.",
  "how_shown": "photoactivatable/fluorescent Ras imaging with acylation-cycle perturbation (PMID 15705808); depalmitoylase identification with localisation readout (PMID 26701913); HPA antibody immunofluorescence",
  "tissue_of_evidence": "non-haematopoietic cell lines throughout; no leukaemic imaging",
  "node": "localization-plasma-membrane-leukaemic-surface-antigens"
}
cell type
{
  "haematopoietic_answer": "ALL of them, uninformatively. HPA Blood Atlas: low immune cell specificity, Detected in ALL immune cell types; RNA tissue specificity low. NRAS expression carries no lineage information, and this entry says so rather than manufacturing a cell-type claim.",
  "what_is_cell-type-specific_is_the_MUTATION_not_the_gene": "The informative unit for NRAS in this disease is the mutant clone and its size - see this entry's cooperating_lesion_profile - not the expression of the wild-type gene. A cell-type field on wild-type NRAS would be a category error of the same family as writing a cell type for a loading control.",
  "paralogue_discipline": "The collision registry's HRAS1 row applies: NRAS (P01111), HRAS (P01112) and KRAS (P01116) are ALL exactly 189 aa and share hotspot codons, HRAS1 is NRAS's only listed synonym, and verify-accessions.py offers zero protection. The acylation-cycle papers cited here name N-Ras explicitly in their titles; nothing may be carried over from an H-Ras or K-Ras result, and the isoforms differ precisely in their C-terminal hypervariable region, which is the part that decides localisation."
}
cell state
{
  "state": "RAS-PATHWAY-ACTIVE, and the graph already holds comut-ras-pathway-kmt2a-r for the pathway-level claim. The localisation-specific state is ACYLATION status: a given NRAS molecule is either palmitoylated and at the plasma membrane or depalmitoylated and at the Golgi, and the ratio is set enzymatically rather than transcriptionally.",
  "unmeasured": "No study has measured NRAS palmitoylation state, or the plasma-membrane-to-Golgi ratio, in any leukaemic cell, let alone a KMT2A-rearranged one."
}
localization bounded absence
{
  "date": "2026-09-21",
  "tool": "NCBI E-utilities esearch, db=pubmed",
  "shape": "gene clause AND localisation clause AND KMT2A/MLL-fusion clause, one [tiab] tag per term (brief rule 94), every cytogenetic term quoted (brief rule 95)",
  "querytranslation_fusion_clause": "\"KMT2A\"[Title/Abstract] OR \"MLL\"[Title/Abstract] OR \"MLL1\"[Title/Abstract] OR \"MLL-AF4\"[Title/Abstract] OR \"MLL-AF9\"[Title/Abstract] OR \"MLL-ENL\"[Title/Abstract] OR \"KMT2A-rearranged\"[Title/Abstract] OR \"MLL-rearranged\"[Title/Abstract] OR \"t 4 11\"[Title/Abstract] OR \"t 9 11\"[Title/Abstract] OR \"MLL fusion\"[Title/Abstract]",
  "count": 2,
  "controls": {
    "fusion_clause_alone": 8743,
    "gene_clause_alone": 8627
  },
  "reading": "Counted in TITLES AND ABSTRACTS only; supporting figures were not searched (brief rule 80).",
  "title_check": "2 records, 0 of them localisation: PMID 21148331 and 15994285 are AML prognosis papers."
}
How this could bite us back
  • If a resistant NRAS clone exists at 0.1% before any therapy, then any agent it resists - and FLT3 D835H clones were found the same way - will select it. This is the mechanism by which a targeted agent converts an invisible clone into the disease.
  • MEK inhibition cannot be escalated in an infant the way it might be in an adult, and a subclonal NRAS target means the achievable exposure is being spent on a minority of cells.
  • An NRAS-wild-type result at standard sequencing depth is not an NRAS-wild-type patient: five of 27 patients in the duplex series had only low-frequency PI3K/RAS mutations. Stratifying a trial on a standard-depth genotype would put those patients in the wrong arm.
  • The published NRAS resistance clones carried G13D and G12S. These are not the alleles that allele-selective RAS inhibitors were built for, so the current generation of mutant-selective RAS chemistry does not reach them.
Still open
  1. Is the 0.06-0.34% NRAS clone present before birth, alongside the fusion, or acquired after? The fusion is a prenatal event in this disease; nobody has looked for these variants on neonatal blood spots at duplex depth.
  2. Does the pre-existing resistant clone predict relapse prospectively? The published observation is retrospective, on six paired samples, and is exactly the kind of result that shrinks when tested forward.
  3. Are NRAS and KRAS mutations in the same patient in the same cell or in different subclones? Bulk sequencing cannot tell, and the 6-of-8 cases where each of two pathway mutations was below 30% MAF suggests different cells.
  4. Does a duplex-depth NRAS clone respond to MEK inhibition differently from a clonal one? A 0.1% clone under selection is a different pharmacological problem from a 40% clone.
  5. NRAS's address is set by a reversible palmitoylation cycle, not by prenylation alone (PMIDs 15705808, 26701913), and nobody has measured that cycle in a leukaemic cell. Does a KMT2A-rearranged blast carrying a subclonal NRAS mutation differ in the plasma-membrane-to-Golgi ratio of NRAS - and would a depalmitoylase- or zDHHC-directed agent reach a clone that farnesyltransferase inhibitors never did?
Every source behind this entry (9)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In 22 whole-genome-sequenced infant MLL-rearranged ALL cases, NRAS was recurrently mutated in 2 cases against KRAS in 4; in 20 non-infant MLL-rearranged leukaemias NRAS was mutated in 3 cases and the kinase/PI3K/RAS pathway mutations were mostly in the major clone, in contrast to infants where they were mostly in minor clones (P < 0.0001).

    The landscape of somatic mutations in infant MLL-rearranged acute lymphoblastic leukemias. · Andersson AK, Ma J, Wang J, et al. · Nature Genetics 2015
    PMID 25730765DOI
    What this evidence cannot show

    Small numbers on both sides of the infant/non-infant comparison (22 and 20 cases), and the non-infant arm mixes ALL and AML. The clonality contrast is reported as a P value on all somatic exonic mutations, not on NRAS alone, so the gene-specific version of the statement is an inference from a category-level result. Full text read via efetch db=pmc (PMC4553269).

    Evidence that argues against this (1)
    • The clonality contrast depends on the detection floor of the assay used on each arm. A duplex-sequencing study of the same disease found that 62% of cancer-associated mutations sit below standard NGS resolution, so a 'major clone versus minor clone' comparison made at 39x compares two truncated distributions. Pilheden M et al. 2022
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In 124 infant B-ALL cases from Interfant-99/06, recurrent mutations were found exclusively in K-RAS and N-RAS, were subclonal, and were frequently lost at relapse.

    PMID 30679323DOI
    What this evidence cannot show

    Abstract read for this entry. 'Exclusively' is a statement about which genes recurred, not about which genes were mutated at all - the same abstract reports a larger number of non-recurrent, non-silent mutations. The cohort is infant B-ALL, overwhelmingly t(4;11).

    Evidence that argues against this (1)
    • Five patients in a duplex-sequenced cohort had only low-frequency PI3K/RAS mutations with no higher-frequency signalling mutation, so cohorts scored at standard NGS depth will misclassify some RAS-mutant patients as wild-type and the recurrence structure is depth-dependent. Pilheden M et al. 2022
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Targeted duplex sequencing of 20 infants and 7 children with KMT2A-rearranged ALL found that 62% of identified cancer-associated mutations were below the resolution of standard NGS, that only 33 of 87 exceeded 2% cellular prevalence (31 of those 33 in PI3K/RAS genes), that five patients had only low-frequency PI3K/RAS mutations, and that drug-resistant clones carrying NRAS G13D or G12S, or FLT3 D835H, comprising 0.06% to 0.34% of diagnostic cells, expanded at relapse.

    PMID 36204688DOI
    What this evidence cannot show

    27 patients, 5 longitudinal and 6 paired relapse samples - a small series, and a targeted panel, so nothing in it bounds total mutation burden. A variant at 0.1% cellular prevalence cannot be assigned to a leukaemic rather than a normal haematopoietic cell by sequencing alone; the inference that these are leukaemic subclones rests on their expansion at relapse. Abstract read; full text not retrieved for this entry.

    Evidence that argues against this (1)
    • The same paper reports infants having FEWER cancer-associated mutations than the older children, so deeper sequencing did not abolish the low-mutation-burden phenotype of infant disease; it relocated part of it below the old floor. Pilheden M et al. 2022
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Sanger sequencing of NRAS, KRAS and BRAF in 109 infant ALL patients found RAS mutations in about 14% overall and about 24% of t(4;11) cases, and RAS mutation was an independent predictor of poor outcome (HR 3.194, 95% CI 1.211-8.429, P = 0.019).

    Frequencies and prognostic impact of RAS mutations in MLL-rearranged acute lymphoblastic leukemia in infants. · Driessen EM, van Roon EH, Spijkers-Hagelstein JA, et al. · Haematologica 2013
    PMID 23403319DOI
    What this evidence cannot show

    The paper reports NRAS and KRAS together as 'RAS mutations'; a gene-specific NRAS hazard ratio is not available from the abstract. Sanger sequencing has a variant-allele-fraction floor around 15-20%, so the cohort is blind to the subclonal fraction and the hazard ratio is therefore a hazard ratio for CLONAL RAS mutation, which is the more informative reading of it.

    Evidence that argues against this (1)
    • A 488-patient paediatric B-ALL cohort found RAS-pathway alterations as a class NOT significantly associated with inferior OS or EFS, with the adverse signal concentrated instead in FLT3/NF1 co-mutation. Chen X et al. 2026
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In 488 paediatric B-ALL cases from the CCCG-ALL-2015 cohort profiled by whole-transcriptome sequencing, 192 (39.3%) carried at least one RAS-pathway alteration, NRAS and KRAS mutations were correlated (phi = 0.19, P < 0.0001) as were NRAS and PTPN11 (phi = 0.138, P = 0.0042), and RAS-pathway alterations as a class were not significantly associated with inferior overall or event-free survival.

    PMID 42440143DOI
    What this evidence cannot show

    Paediatric B-ALL as a whole, not infant KMT2A-rearranged disease, and a single-centre retrospective cohort. Mutations were called from whole-TRANSCRIPTOME sequencing, which detects only expressed alleles and has an allele-fraction floor set by expression level, so both the 39.3% and the co-occurrence coefficients are RNA-level measurements. The phi coefficients are small even where the P values are convincing. Abstract read; full text not retrieved.

    Evidence that argues against this (1)
    • An Interfant cohort found RAS mutation to be an independent predictor of poor outcome in KMT2A-rearranged infant ALL specifically (HR 3.194), so the null result for the class in unselected paediatric B-ALL should not be transferred to this disease. Driessen EM et al. 2013
  6. E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    UniProt P01111 records NRAS at the cell membrane and the Golgi apparatus membrane, both as a lipid anchor, and states that it shuttles between the two; HPA immunofluorescence gives main location Plasma membrane with Cytosol additional at reliability Supported.

    What this evidence cannot show

    Database comment on one day, from two publications. A 'Cytosol' additional call in HPA for a lipid-anchored protein most likely reports the unprocessed or depalmitoylated pool rather than a genuine free cytosolic species. No haematopoietic system in the coded citations.

    Evidence that argues against this (1)
    • 'Plasma membrane' is a majority statement rather than a location, and treating it as the answer hides the only tractable thing about this protein. The same record carries a Golgi pool and HPA adds a Cytosol call; the split between them is set enzymatically and is pharmacologically movable, which is now being exploited by engineering NRas palmitoylation to capture zDHHC inhibitors. A static compartment field for NRAS records the equilibrium and discards the variable. Cell Chem Biol 2026
  7. E7established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    A continuous acylation cycle - palmitoylation trapping N-Ras at the plasma membrane and depalmitoylation returning it to the Golgi - maintains the plasma-membrane pool and governs where palmitoylated Ras isoforms signal from.

    PMID 15705808
    What this evidence cannot show

    Non-haematopoietic cells, 2005, and the paper covers palmitoylated Ras isoforms as a class - N-Ras and H-Ras differ in the number of palmitoylation sites, so the quantitative behaviour is not identical between them. Read at title-and-abstract level. Nothing in it is leukaemic.

    Evidence that argues against this (1)
    • The result is about palmitoylated Ras isoforms AS A CLASS and does not transfer to NRAS quantitatively, because the isoforms differ precisely in the C-terminal region that drives the cycle: HRAS is dually palmitoylated, NRAS singly, and KRAS4B not at all. The therapeutic consequence shows the difference is real - farnesylation-directed therapy remains a live strategy in HRAS-mutant disease while it failed against NRAS - so a class-level acylation model cannot be used to predict NRAS behaviour without isoform-specific data. Oncogene 2022
  8. E8strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The ABHD17 family are depalmitoylases that regulate N-Ras palmitate turnover and thereby its subcellular localisation, identifying an enzymatic control point over the address rather than over the abundance of NRAS.

    PMID 26701913
    What this evidence cannot show

    Names N-Ras explicitly, which matters given the 189-aa/hotspot-shared RAS collision. Non-haematopoietic systems; no leukaemia and no KMT2A fusion. Abstract-level reading. Whether ABHD17-directed chemistry is tractable is not addressed here.

    Evidence that argues against this (1)
    • Identifying a depalmitoylase does not make it the exploitable node, and the field has largely moved the other way: current chemistry targets the FORWARD enzymes, with NRas palmitoylation engineered to capture zDHHC inhibitors rather than ABHD17 inhibitors being pursued. A reversible cycle has two sides and blocking removal would be expected to trap NRAS at the plasma membrane - the opposite of the wanted effect - so this entry's implication that ABHD17 is a handle needs to be read as one of two possible directions, not as the direction. Cell Chem Biol 2026
  9. E9strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    Human Protein Atlas v25.1 records NRAS as detected in all immune cell types with low immune cell specificity, so expression of the wild-type gene carries no haematopoietic lineage information.

    What this evidence cannot show

    RNA in normal blood. This is a deliberately negative cell-type record: the informative unit for NRAS in this disease is the mutant clone, not wild-type expression.

    Evidence that argues against this (1)
    • The deliberately negative cell-type record can itself be over-read. A flat RNA profile does not mean NRAS is uninformative by cell type, because the acylation machinery that sets its address is cell-type-regulated: identical NRAS mRNA in two cells can mean different plasma-membrane-to-Golgi ratios if their depalmitoylase or zDHHC complements differ. So 'no lineage information' is true of expression and untested for localisation, and this entry should not be read as saying NRAS has no cell-type biology. Elife 2015
Connected to