Cell fusion may restore growth control through resident genes in aging epithelial tissue

The train of thought that ends in this hypothesis. Each stage is the reason the next exists. The master question narrows to a goal, the goal to an unknown nobody has closed, the unknown to the hypothesis proposed here. Every step below says what it rests on and what carries it.

Two established results predict opposite outcomes, and both cannot be right.

Does suppressing expanding cell families in aging human linings promote invasion or improve repeated repair?

The engine's own statement of the hypothesis, in full.

HERETICAL, conditional: In a subset of aging human epithelial mosaics, rare somatic fusion between a nonmalignant resident and an emerging invasive lineage may create viable hybrid descendants whose retained resident-derived growth-restraining allele or alleles restore invader growth control. The central mechanism is functional genome complementation within an invader-derived descendant; hybrid ancestry alone, generic polyploid stress or checkpoint activation would not establish it. If quantitatively consequential, removing fusion-competent residents could remove this component of protection even when fusion-incompetent replacement preserves successful daughter replacement, mature output and occupation. Nonlethal suppression of resident division could preserve the complementation component during a measured interval in which encounter rates, actual fusion incidence, functional allele expression and barrier competence remain adequate. Existing arrested hybrids need not resume growth when formation of new hybrids is blocked. Hybrids may instead remain invasive or later escape through chromosome loss; fusion alone never classifies a resident as beneficial. A contribution to limiting dangerous lineage expansion is possible, while SPV_9, anatomical invasion and whole-tissue function remain separate outcomes. Ordinary turnover competition can operate alongside this route.

Every other hypothesis the engine wrote for the same gap, and the observation that would separate the two.

• Rival 01 of 03

  Resident daughter cells may restrain epithelial invaders by replacing their renewing positions
  What would separate them

  Resident daughter cells may restrain epithelial invaders by replacing their renewing positions predicts: Proposed common-condition test: in a sterile immune-free older-donor epithelial mosaic of one site, track a complete resident–invader boundary fate ledger: division, daughter retention and differentiation, migration, extrusion direction, death, vacancies and actual replacement of viable invader renewing positions. Keep invader genotype, initial lineage composition, matrix, baseline nutrients and functional permeability comparable and measure local contact geometry, density, cell size, tension proxies and signaling. First test whether gross J_R predicts directly observed h_RI at all. Then reversibly alter resident output with a resident-autonomous perturbation and rescue during the measured window before mature function or coverage diverges, with direct invader-exposure controls. If the mechanism operates, a reduction in successful h_RI should reduce position-loss events and increase absolute renewing-invader growth by the event-accounted amount when other rates are unchanged; restoring h_RI with competent fusion-incompetent cells should restore that component of restraint. Equal J_R alone does not establish equivalent replacements, and no mutation-count-independent invasion claim is made without equivalent contacts, hazards and function. Estimate parameters from mutually exclusive observed fates, predict held-out episodes, and vary geometry explicitly rather than absorbing every result into kappa_R. Gross production changes with no h_RI change and no growth effect favor a contact/occupancy explanation over a transferable gross-flux law. Coupled mechanical and replacement effects support spatial turnover competition and limit an independent constant-coefficient model. Measure mature output and next-cycle reserve independently; a lower g_I cannot compensate for their deterioration. If toxin or feeding interruption removes an ablation-associated excess while h_RI stays low, this limits flux loss as a sufficient or dominant cause only when invader growth capacity is preserved. Nutrients can increase lambda_I or survival enough to overcome high h_RI without disproving displacement. Cross matrix strata that preserve both lineages and allow events; matrix effects may alter h_RI or other rates and can coexist, while a selective h_RI null bounds this contribution. Report SPV_9 with denominator growth and anatomical crossing separately. Additional required discriminators and bounded interpretations: Before observing held-out net growth, estimate time/geometry-conditioned rate relationships on other repairs, then integrate their predicted local trajectories and propagate event-recovery, parameter and ambiguous-fate uncertainty. Compare predicted absolute I and separate tissue clone fractions with observations. A failed stationary approximation limits that reduction, not the possibility of displacement; post hoc reconstruction from the same outcome episode is bookkeeping, not predictive transfer. Follow P_R(t), mature daughter production, D_R survival/lifespan, tissue-specific mature function, growth control and reserve available at the next repair jointly. Integrate mature inflow and loss over time; a lower renewing-invader growth rate or higher J_R does not establish beneficial aging renewal or adequate specialization. Track divisions, daughter states, nondivision loss and mature survival separately; audit migration/dedifferentiation and include their net flows or restrict the model to a justified closed interval. Estimate time/geometry-conditioned rates and prospectively integrate held-out repair trajectories with uncertainty, preserving specialized function as an independent outcome. Directly count gross progenitor daughters per measured area/time and separately follow which reach and causally replace viable invader positions. Record changing area and boundaries; compare J_R with the event-resolved hazard in held-out geometry/time strata, allowing no relation or changing kappa_R. Freeze training-derived time/geometry relationships and predictive uncertainty before held-out episodes, including reciprocal replacement and repair-stage changes; use observable starting state and prespecified inputs. Recounting all held-out fates afterward checks accounting only. If no useful stable relation survives, retain event-caused displacement conditionally and reject or bound that transportable reduction, not replace its mechanism. Prespecify invader-derived hybrid membership and count each viable descendant once in a lineage numerator, track renewing and differentiated compartments and denominator dynamics, and report SPV_9 per year separately from absolute growth and crossing. Check observed count consistency and uncertainty; changing fractions or mutation counts alone cannot establish malignancy or whole-human aging benefit. Compare tracked fates, absolute growth and post-colonization balance in matched normal and tumor contexts with recorded genotypes and perturbation/rescue. Independently test interlineage competition and serial mature/barrier function; a fitness sign reversal alone neither validates any native candidate nor predicts older-human cancer prevention. In the same older-donor epithelial model compare retained versus removed residents with matched initial wounding and adequate functional replacement, and separately vary immune presence when feasible. Track resident–invader contacts, direct losses, basal extrusion, matrix and mature function. Resident protection is supported only as a measured contribution after removal-induced injury alternatives are tested; immune-free behavior alone does not establish in-vivo immunity irrelevance. Compare genotype-resolved survival and fate under recorded inflammatory exposure versus comparable noninflamed conditions, with functional perturbation/rescue and serial mature/barrier outcomes. Prospectively track cancer outcomes separately. Any colitis-context resistance prediction does not imply an age-only benefit or make a resident universally protective. Factorially vary defined matrix chemistry/architecture and selective signaling perturbation/rescue within a recorded genotype and matched repair context; measure actual crosslinks, focal signaling, basement-membrane crossing, mutation acquisition and wound strength separately. A matrix contribution can coexist with all four native mechanisms and does not establish absence of new mutations merely because none were intended. Use time-resolved boundary imaging plus selective resident output/incorporation perturbation and rescue, comparing the ordering of viable position occupancy, incorporation and invader exit. Prespecify uncertain-order cases and bound their influence; preserved area or gross J_R alone cannot identify replacement causation. Calibrate fate observation and missed events; report growth bounds under plausible assignments of ambiguous events without allocating them to fit the equation. Track displaced descendants as viable differentiated, basally displaced, dead or unresolved separately: loss from the renewing compartment is not automatically loss of the clone or anatomical safety. Current finding-specific qualification: The exponential compartment trajectory applies to a deterministic or justified conditional-mean reduction with homogeneous prescribed rates and represented flows; it is not a realized rare-clone trajectory or an extinction law. For random history-dependent rates, dE[I_s]/dt=E[g_s I_s]+E[net flows], and E[g_s I_s] generally differs from E[g_s]E[I_s]. A product-of-means closure requires independent justification. Prespecify and validate the closure on training repairs or integrate an explicit history-dependent stochastic model; propagate event, rate and demographic variability into held-out expected counts and outcome distributions. Report realized counts and extinction separately, without replacing the already defined time-local hazard by the pooled interval estimator. Compare equally prespecified contact/occupancy models. Match baseline mechanics; measure mechanics caused by daughter incorporation as part of its total displacement effect instead of clamping that mediator. A failed closure or proportional hazard reduction limits that reduction, not event-caused displacement.

• Rival 02 of 03

  Loss of mucosal barriers may let microbial toxins create mutations that drive invasion
  What would separate them

  Loss of mucosal barriers may let microbial toxins create mutations that drive invasion predicts: Proposed common-condition test: use a defined older-donor intestinal epithelial mosaic supporting documented luminal microbial contact and repair, with banked, deeply characterized ancestral lineages split across conditions. Cross retention, viability-preserving suppression and controlled small-area removal with verified toxin-competent, toxin-deficient and complemented producer conditions plus sterile controls. The matched reagent pair is an experimental target: verify microbial growth, metabolism, adhesion, community composition and proximity; equal bacterial counts are not equal delivered dose. Match initial deletion, non-toxin injury, starting producer burden, baseline nutrients, initial corpse material and matrix; downstream toxin-induced corpse release is not matched away in the total-effect arm. Measure functional permeability and local toxin activity/contact over time and do not force downstream toxin-mediated injury equal, because it lies on the proposed causal path. First establish whether removal actually increases delivered exposure and DNA adducts/damage in surviving renewal cells. If it does, follow barcodes, repeated samples and descendant phylogenies to nominate sequence changes acquired after exposure and before phenotype, reporting deep-baseline detection limits rather than claiming every rare pre-existing subclone was absent. A new mutational signature or passenger plus expansion is insufficient. After toxin withdrawal, compare descendants in a shared uninjured niche and matrix; reconstruct a nominated acquired lesion in the appropriate ancestral background or correct it in the descendant, with matched genetic controls, to test causal invasive behavior and possible cooperation with pre-existing drivers. Under the hypothesis, specific toxin interruption reduces the acquired-lesion-mediated component, and restoration of verified delivered exposure is predicted to restore its population-level contribution under a viable exposure regime, not the identical lesion or descendant in every replicate. Toxin dependence alone can also reflect selective killing or growth of pre-existing invasive lineages; ancestry and lesion causality distinguish these routes. Sterile or toxin-negative arms should lack this toxin-mediated new-lesion component, while other release mechanisms can persist. A finite negative result bounds the tested dose/time and detectable lineage frequency, not all late events. Measure absolute descendant production, clone fractions with denominator, mature barrier recovery and basement-membrane crossing separately. Matrix-dependent invasion without required new causal lesions supports a matrix contribution and can coexist with genotoxicity. Additional required discriminators and bounded interpretations: Measure local delivered dose, activity/contact, adducts/damage, cell death/arrest and surviving descendant lesions in banked ancestral lines under matched toxin-deficient and complemented conditions. Bound pre-existing ancestry and require separate lesion-function evidence before predicting invasive descendants; current microbiome composition cannot establish past exposure. Compare tracked fates, absolute growth and post-colonization balance in matched normal and tumor contexts with recorded genotypes and perturbation/rescue. Independently test interlineage competition and serial mature/barrier function; a fitness sign reversal alone neither validates any native candidate nor predicts older-human cancer prevention. In the same older-donor epithelial model compare retained versus removed residents with matched initial wounding and adequate functional replacement, and separately vary immune presence when feasible. Track resident–invader contacts, direct losses, basal extrusion, matrix and mature function. Resident protection is supported only as a measured contribution after removal-induced injury alternatives are tested; immune-free behavior alone does not establish in-vivo immunity irrelevance. Compare genotype-resolved survival and fate under recorded inflammatory exposure versus comparable noninflamed conditions, with functional perturbation/rescue and serial mature/barrier outcomes. Prospectively track cancer outcomes separately. Any colitis-context resistance prediction does not imply an age-only benefit or make a resident universally protective. Factorially vary defined matrix chemistry/architecture and selective signaling perturbation/rescue within a recorded genotype and matched repair context; measure actual crosslinks, focal signaling, basement-membrane crossing, mutation acquisition and wound strength separately. A matrix contribution can coexist with all four native mechanisms and does not establish absence of new mutations merely because none were intended. Compare retention, suppression and the specified removal procedure using delivered toxin activity/contact, functional barrier measures and surviving-renewal-cell damage. Proceed to the lesion-mediated prediction only if that first link is supported in the proposed model; absence under rapid functional replacement bounds that procedure, not all epithelial removal. Preserve banked ancestral lines, baseline detection bounds, time-resolved ancestry and post-exposure lesion timing. Test nominated new changes in the relevant ancestral/descendant genetic background, with existing drivers characterized, and require a causal contribution to sustained anatomical invasion after washout in a justified common niche rather than signature enrichment or toxin dependence alone. Across independent ancestral replicates compare exposure-linked rates and distributions of surviving causal-lesion descendants, with uncertainty, delivered dose, cell loss/arrest and adequate follow-up recorded. Restoration means recovery of the conditional distributional contribution within a justified exposure regime; a finite unmatched stochastic realization is not falsification, and no universal monotonic dose response is asserted. Report both total-effect and pulse-clamped estimates with their distinct intervention conditions; track all initial lineages, death and surviving descendants, then report survivor-conditional quantities as such. Preserve the toxin-to-death-to-feeding chain in total-effect comparisons and use ancestral matching rather than treating survivor selection as randomization. Where indicated, test justified lesion sets and characterized cooperating state/background with matched genetic controls. Retain the exposure-to-ancestry chain and test sustained basement-membrane crossing in a shared niche justified for the original tissue. Invasion enabled only by an artificial permissive matrix does not establish consequential invasion caused by removal in aging tissue. Current finding-specific qualification: Toxin restoration remains a conditional distributional prediction, with distinct estimands per initial lineage, per exposure and conditional on survival. A pulse-clamped contrast is a separate controlled effect only if the clamp is feasible and does not independently alter delivered toxin, susceptibility or relevant repair functions; feasibility is not established. Prespecify follow-up and report the probability of at least one causal-lesion descendant per starting lineage, absolute causal-lesion descendant counts per initial lineage, and any exposure-normalized rate with its exposure denominator and uncertainty; separately label survivor-conditional quantities and follow all initial lineages including deaths/arrests. Validate any pulse clamp for exposure and susceptibility; if validation fails, omit its isolated causal interpretation while retaining the total-effect arm with toxin-induced injury, death and feeding intact. Validate producer deletion/complementation ecology and delivery; toxin dependence alone still does not distinguish new-lesion causality from selection.

• Rival 03 of 03

  Dead resident cells may feed invasive lineages during epithelial repair
  What would separate them

  Dead resident cells may feed invasive lineages during epithelial repair predicts: Proposed common-condition test: in a defined older-donor epithelial mosaic with a barcoded pre-existing invasive lineage, initially exclude microbes and immune cells. Compare resident retention, viability-preserving suppression, killing with retained debris and physical extraction, recording local coverage, successful replacement hazard, mature function, death mode, nutrient delivery, matrix and non-target injury. Prelabel resident biomass, wash and chase, then measure the amount, chemical species and kinetics released, cleared and reaching invaders. Trace label through extracellular species into purified newly synthesized invader macromolecules and, for survival/maintenance, quantify catabolic products and energetic use; exclude residual medium/free label, adherent corpses, undegraded engulfed material and acellular carryover. Add macrophages only in separate arms to distinguish direct from indirect recycling. Quantify absolute births, deaths, net newly produced biomass and invasive events. Use carbon/nitrogen or substrate-relevant mass balance, with medium inputs, recovered corpse material and metabolic losses, to ask whether accessible assimilated material and its measured yield plausibly account for extra output; a limiting trace substrate need not supply most biomass, but its necessity and quantitative capacity must be shown. Test measured resident-equivalent concentration/time courses under baseline-supporting, physiologically defensible nutrient/perfusion and clearance conditions; dependence manufactured by starvation does not establish an ordinary-human route. After identifying a candidate limiting substrate, use orthogonal invader-specific uptake/conversion perturbations with target engagement and demonstrated loss of relevant flux, rescue the pathway, and compare equally injured invaders supplied with non-corpse nutrients. Baseline fitness alone cannot exclude a general repair-stress or signaling effect, and one redundant transporter block is not a complete salvage interruption. Under the hypothesis, selective loss of material flux reduces the extra production/survival component and its restoration rescues it. Fraction sufficiency at measured kinetics is supplementary: separately interrupt signaling versus metabolic conversion where feasible and test alternative nutrient rescue to distinguish instructions from feeding. If signaling drives extra growth despite adequately verified loss of material flux, the feeding contribution is bounded; if selective separation is impossible, label plus growth does not establish nutrition. High-flux or fusion-competent resident add-back can consume the pulse or suppress fed invaders, so restored restraint does not disprove feeding. Measure actual uptake and birth/loss terms in such factorial arms; matrix may separately determine crossing. Null invasion alone does not exclude nutrition, and null single-step perturbation alone does not exclude redundant salvage. Additional required discriminators and bounded interpretations: For the survival branch extend substrate-relevant mass balance to extracellular metabolites, intracellular flux, catabolic products and energetic/maintenance use, while separating contamination and indirect macrophage recycling. Couple measured flux to absolute survival/death and selective loss/rescue; absence of retained biomass label alone cannot exclude catabolic feeding, and catabolic label alone cannot establish its causal benefit. Bound accessible amount, uptake rate and plausible output supported using the measured requirement/yield and uncertainty; test graded physiologically defensible availability with alternative nutrient rescue and pathway-selective loss/rescue when feasible. Do not equate bulk isotope enrichment with causal contribution or claim a numerical capacity absent measurements. Compare tracked fates, absolute growth and post-colonization balance in matched normal and tumor contexts with recorded genotypes and perturbation/rescue. Independently test interlineage competition and serial mature/barrier function; a fitness sign reversal alone neither validates any native candidate nor predicts older-human cancer prevention. In the same older-donor epithelial model compare retained versus removed residents with matched initial wounding and adequate functional replacement, and separately vary immune presence when feasible. Track resident–invader contacts, direct losses, basal extrusion, matrix and mature function. Resident protection is supported only as a measured contribution after removal-induced injury alternatives are tested; immune-free behavior alone does not establish in-vivo immunity irrelevance. Compare genotype-resolved survival and fate under recorded inflammatory exposure versus comparable noninflamed conditions, with functional perturbation/rescue and serial mature/barrier outcomes. Prospectively track cancer outcomes separately. Any colitis-context resistance prediction does not imply an age-only benefit or make a resident universally protective. Factorially vary defined matrix chemistry/architecture and selective signaling perturbation/rescue within a recorded genotype and matched repair context; measure actual crosslinks, focal signaling, basement-membrane crossing, mutation acquisition and wound strength separately. A matrix contribution can coexist with all four native mechanisms and does not establish absence of new mutations merely because none were intended. Quantify release, delivery, uptake and use versus viable/suppressed-resident secretion, and test whether a limiting material input causes extra survival or production after controlling independent instructive signaling while allowing downstream nutrient-state signaling to vary. Equivalent dead mass from another lineage predicts equivalence only when species, composition, death mode, clearance and local availability are matched. Demonstrate loss of the relevant material flux, intact matched-injury responses and alternative-nutrition growth, and orthogonal pathway-specific rescue. If redundancy or inseparable signaling prevents these controls, report transfer plus association and bounds, not proven nutritional causation; a single null transporter perturbation does not reject the mechanism. Measure local release, availability, uptake and output over an explicitly justified delivery/clearance range anchored to the chosen tissue once such data exist. Report the range as unvalidated until measured. An immune-free arm isolates direct transfer; separately add phagocyte clearance/recycling and record net effects before making any in-vivo inference. Current finding-specific qualification: Material feeding can cause downstream nutrient-state signaling. Preserving independent instructive signaling in the nutrition contrast must not mean clamping signaling caused by restored limiting material or energy. The rival is instruction independent of that limiting material function, not every signaling response to nutrition. Where selective separation is feasible, compare material-use loss and orthogonal pathway or suitable nutrient restoration while preserving independent instructive exposure and matched-injury/alternative-nutrition competence, allowing downstream nutrient-state responses to vary. Attribute assimilative and catabolic/maintenance flux to viable invaders with the existing sterile, acellular and separate-phagocyte controls. If material and instructive roles cannot be separated, report transfer, association and substrate-specific quantitative bounds rather than proven nutritional causation.

The engine is asked to say what its hypothesis would overturn and what would surprise a specialist. This is its answer.

Which of the figures above have a study behind them, which are the engine's own, and what it would take to refute the hypothesis. This audit never judges the idea.