Pan-essentiality: checked for all eight, and cleared — but a shallow dependency is its own problem
also called common essential check · pan-cancer Chronos FA genes
A gene that every cell needs is useless as a cancer drug target, because blocking it would harm healthy cells just as much. All eight Fanconi genes behind this finding were checked against every cell line ever screened, and none of them is in that category. The catch is the opposite one: the effect is small enough that there may not be much of a window either way.
The same thing, for a scientist
This project's earlier scans were contaminated by common essentials — ribosomal and core-machinery genes that scored as KMT2A-selective only because the rearranged panel was globally more sensitive — so the pan-essentiality filter is mandatory before any selectivity claim. Computed across every screened line in the Open Targets DepMap slice rather than only leukaemia, all eight headline genes sit between Chronos -0.09 and -0.33 with 4.6% to 16.6% of lines below -0.5, against -0.815 and 72.1% for ACTB [E1]. None is a common essential. That clears the necessary condition and nothing more. Two problems remain and they point in opposite directions. **The window is narrow:** a KMT2A-selective difference of -0.23 to -0.44 on a pan-cancer baseline of -0.09 to -0.33 means the absolute requirement in the index lines is still modest, roughly -0.5 to -0.7, which is far from the -2.1 MEN1 reaches in MOLM-13. **And the baseline is not zero:** a mean Chronos of -0.30 across all lineages means a systemic inhibitor would be exerting real pressure on every proliferating tissue, which for this pathway means the haematopoietic compartment where Fanconi anaemia's phenotype lives.
The numbers
Across every screened cell line in the Open Targets DepMap slice, not only leukaemia 1.
| gene | pan-cancer Chronos mean | % of lines below -0.5 | % below -1.0 | KMT2A-selective d |
|---|---|---|---|---|
| FAAP100 | -0.332 | 14.3% | 0.6% | -0.270 |
| BRIP1 | -0.315 | 16.6% | 0.4% | -0.343 |
| FANCI | -0.299 | 16.4% | 0.7% | -0.361 |
| FANCF | -0.298 | 15.6% | 0.6% | -0.437 |
| FANCL | -0.260 | 11.3% | 0.3% | -0.395 |
| FANCA | -0.259 | 14.2% | 0.4% | -0.357 |
| UBE2T | -0.228 | 9.1% | 0.3% | -0.368 |
| FANCG | -0.091 | 4.6% | 0.1% | -0.314 |
| MEN1 (comparator) | -0.363 | 29.3% | -0.863 | |
| ACTB (contaminant shape) | -0.815 | 72.1% | -0.179 |
ACTB is the instructive row. It has the deepest pan-cancer requirement of anything in this table and the smallest KMT2A-selective difference, and its selectivity evaporates in the lymphoid half (+0.066). That is what a global-sensitivity contaminant looks like, and none of the eight resembles it.
Why 'not pan-essential' is a weaker result than it sounds
Three things it does not establish.
It is a necessary, not sufficient, condition. A gene can be dispensable in every lineage and still not be selectively required in KMT2A-rearranged cells. Pan-essentiality screening removes a class of false positives; it adds no positive evidence.
It says nothing about normal tissue. DepMap screens cancer cell lines. A Chronos near zero across 1,200 tumour lines does not predict tolerance in a patient's haematopoietic stem cells, and for this pathway there is direct human evidence that it will not be tolerated: biallelic loss of any of these genes causes Fanconi anaemia, with progressive bone-marrow failure. The pathway's normal-tissue requirement is one of the best-documented in human genetics and it is not visible in this filter at all.
The therapeutic window is the difference, not the ratio. In the index lines the absolute requirement is roughly the pan-cancer baseline plus the selective difference, about -0.5 to -0.7. MEN1 reaches -2.1 in MOLM-13. A drug would have to exploit a fitness difference several times smaller than the one menin inhibitors exploit, against a target class with a named marrow toxicity.
The one gene where the filter says something specific
FANCG has a pan-cancer Chronos of -0.091 with only 4.6% of lines below -0.5 — essentially dispensable everywhere — yet a KMT2A-selective difference of -0.314. That is the cleanest selectivity ratio in the set, and on a naive target-triage it would rank first.
It is also the gene whose lineage behaviour is the odd one out: FANCG is the only one of the eight more depleted in the myeloid half than the lymphoid, which is the artefact-prone direction for this project. And a near-zero baseline makes the difference statistic unstable, because a small absolute shift produces a large apparent selectivity. The two observations should be held together: FANCG has the best ratio and the least robust evidence, and those are related facts.
| kind | essential-function |
|---|---|
| severity | tested and cleared as a confound; replaced by a narrow-window concern and a normal-tissue concern |
| mitigation | treat pan-essentiality as a necessary filter only; for this target class the binding constraint is normal haematopoietic tissue, which DepMap cannot measure |
affects
[ "gene-fancf", "gene-fanci", "gene-fanca", "gene-brip1", "gene-fancl", "gene-fancg", "gene-ube2t", "gene-faap100" ]
pan cancer chronos — the numbers behind the chart
{
"FAAP100": -0.3318,
"BRIP1": -0.3147,
"FANCI": -0.2994,
"FANCF": -0.2975,
"FANCL": -0.2603,
"FANCA": -0.2588,
"UBE2T": -0.2283,
"FANCG": -0.0905,
"MEN1": -0.3631,
"ACTB": -0.8145
}frac lines below minus 0 5 — the numbers behind the chart
{
"FAAP100": 0.143,
"BRIP1": 0.166,
"FANCI": 0.164,
"FANCF": 0.156,
"FANCL": 0.113,
"FANCA": 0.142,
"UBE2T": 0.091,
"FANCG": 0.046,
"MEN1": 0.293,
"ACTB": 0.721
}- The therapeutic window here is several times narrower than the one menin inhibitors exploit, against a target class whose loss-of-function phenotype in humans is bone-marrow failure. That combination is the standard reason a dependency never becomes a drug.
- A pan-cancer Chronos around -0.30 means a systemic inhibitor would press on every proliferating tissue, not only the leukaemia.
- What is the FA-gene requirement in normal human CD34+ haematopoietic stem and progenitor cells, which is the comparator that actually determines the therapeutic window?
- Is FANCG's exceptional selectivity ratio real, or an instability of the difference statistic on a near-zero baseline?
- Does the absolute requirement in the index lines (roughly -0.5 to -0.7) rise in a more physiological model, or fall?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
None of the eight headline FA genes is a common essential: mean Chronos across all screened lines runs -0.091 (FANCG) to -0.332 (FAAP100), with 4.6% to 16.6% of lines below -0.5 and under 1% below -1.0, against -0.815 and 72.1% for ACTB.
Is the Fanconi signal lineage-wide, fusion-specific, and separable from pan-essentiality? Lineage-split, subtype-control and line-label permutation re-test of the FA/BRCA enrichment · Oncofusion knowledge graph, analyses/fa-pathway-lineage-and-fusion-specificity; DepMap Chronos via Open Targets 2026What this evidence cannot show
Computed from the Open Targets slice of DepMap, which does not state its release version, and 'all screened lines' means all lines the API returns for that gene, which varies by gene. This is a computed pan-cancer mean, not DepMap's own curated 'common essential' label.
Evidence that argues against this (1)
- A gene being dispensable across tumour lines says nothing about normal haematopoiesis. Biallelic loss of any of these eight causes Fanconi anaemia with progressive bone-marrow failure in humans, which is the exact tissue and the exact toxicity a leukaemia therapy would encounter, and no cancer-cell-line filter can see it. Quentin S et al. 2011
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
The pan-essentiality filter is necessary in this project specifically, because earlier KMT2A-selectivity scans were contaminated by ribosomal and core-machinery genes that scored only because the rearranged panel is globally more sensitive.
Effect-size versus z re-ranking of the genome-scale lineage-consistency sweep, with a permutation test of the Fanconi/FA-BRCA pathway · Oncofusion knowledge graph, analyses/effect-size-vs-z-reranking; DepMap Chronos via Open Targets 2026What this evidence cannot show
The per-line offset correction used throughout this beat already removes most of that global-sensitivity effect, so the pan-essentiality filter is partly redundant with the normalisation rather than an independent check.
Evidence that argues against this (1)
- ACTB survives the offset correction with a myeloid-half d of -0.319, so the normalisation does not fully remove the contaminant shape and the filter is doing real work. Oncofusion knowledge graph, analyses/fa-pathway-lineage-and-fusion-specificity 2026
tested: none of the eight is pan-essential, but the window is narrow
tested: none of the eight is pan-essential, but the window is narrow
- related to →FANCG — FA core complex; rank 27 of 8,269 for KMT2A-selective dependency, and no KMT2A literature at all
FANCG has the best selectivity ratio (pan-cancer Chronos -0.091) and the least robust evidence — it is the only one of the eight stronger in the myeloid half
the filter that cut 22 candidates to five
the pan-essentiality filter and its blind spot
18 of the 23 largest differences here are pan-essential