Density-dependent growth emerges from Bayesian adaptation of phenotype
Manish Kumar Gupta, Arnab Barua, Haralampos Hatzikirou · 2026
Read Density-dependent growth emerges from Bayesian adaptation of phenotype by Manish Kumar Gupta, Arnab Barua, Haralampos Hatzikirou (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-25918.txt.
Introduction
Manish Kumar Gupta, Arnab Barua, and Haralampos Hatzikirou ask why tumour growth often refuses to follow the classic exponential curve. Real tumours can grow faster or slower depending on how many cells are already present, stall at low density, peak at middle density, and level off before food or space runs out. Standard models impose such density dependence by hand, without explaining where it comes from.
Cells are modelled as Bayesian agents in their framework. Each cell senses signals produced by the whole population and adjusts its coarse-grained phenotype, its current operating mode, toward states that explain those signals well. A mismatch between sensing and the growth optimum grows with population size and bends the growth rate down quadratically. Adding a receptor-ligand decoding model makes the mismatch non-monotonic, yielding a middle-density growth optimum, an Allee survival threshold, and tissue-specific capacity from one mechanism.
Catalog record
| CREATOR | Manish Kumar Gupta, Arnab Barua, Haralampos Hatzikirou |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 112 KB |
| LICENSE | http://creativecommons.org/licenses/by-sa/4.0/ |
| RIGHTS | CC BY SA |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.25918
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Density-dependent growth emerges from Bayesian adaptation of phenotype Manish Kumar Gupta, Arnab Barua, Haralampos Hatzikirou Abstract Classical models often describe early tumor expansion as exponential growth, yet experimental and clinical evidence shows that tumor populations can deviate systematically from this behavior, exhibiting density-dependent proliferation, cooperative low-density growth, intermediate growth optima, and finite upper growth bounds before resource limitation or spatial crowding dominate. These observations raise a common question: why should the per-capita growth rate depend on population size? Here, we propose that sensing mismatch provides a mesoscopic link between environmental change and density-dependent proliferation. We model the cell as a Bayesian adaptive agent whose coarse-grained phenotype evolves on an intrinsic regulatory landscape, while environmental sensing reweights phenotypic states according to how well they account for the extracellular signal statistics generated by the population. In the weak phenotype–signal correlation regime, the stationary phenotype distribution is Gaussian, with its mean displaced from the proliferative optimum by a population-size-dependent baseline information mismatch. This displacement produces a quadratic penalty in the per-capita growth rate. Coupling the framework to a receptor–ligand decoding model, we show that basal readout error and nonlinear receptor saturation make the mismatch non-monotonic in population size. This single structure gives rise to an intermediate proliferation optimum, an Allee survival threshold, a tissue-specific capacity, and superlinear scaling at low density. A phase diagram in the phenotype–signal coupling and readout-error plane partitions growth into regulated, uncontrolled, and arrested regimes. Thus, density-dependent proliferation need not be imposed phenomenologically, but can emerge from cellular sensing and inference. [orcid=0009-0003-9649-1894] [orcid=0000-0002-8210-5978] [orcid=0000-0002-1270-7885] keywords Growth laws,Density-dependent growth,Decision making ,Bayesian learning ,Phenotype adaptation † † credit: Investigation, Conceptualization, Methodology, Algorithm design, Writing – original draft † † credit: Conceptualization, Writing – review \And editing † † credit: Investigation, Conceptualization, Supervision, Writing – review \And editing † † corresponding: Corresponding author 1 Introduction Understanding how cells regulate proliferation in response to population size remains a central problem in developmental biology and cancer research [ 27 , 32 ] . This question is especially important in cancer, where the proliferative behavior of small tumor-cell populations determines whether they collapse, persist, or expand, thereby shaping tumor initiation, recurrence, and metastatic outgrowth [ 28 , 20 , 22 ] . Classical descriptions of early tumor expansion assume exponential growth, in which each cell proliferates independently and the population growth rate is proportional to the number of cells present. This assumption is mathematically convenient and biologically plausible when cells behave as autonomous proliferative units. Several recent experimental and clinical observations indicate that tumor growth deviates substantially from this picture. Tumor metabolic activity and proliferation-related imaging signals have been reported to scale superlinearly with tumor volume across multiple human cancers [ 26 , 4 , 23 ] , while controlled in vitro measurements at low seeding densities show Allee-like behavior, in which the per-capita growth rate increases with population size [ 15 , 24 ] . Together with the existence of a finite carrying capacity and intermediate proliferation optima, these phenomena raise a common question: Why should the per-capita growth rate of a cell population depend on its size, even before classical resource limitation occurs ? To answer this question, researchers have proposed two broad mechanisms. The first is the evolutionary mechanism: Pérez-García and colleagues showed that superlinear tumor growth can arise if larger populations accumulate more aggressive subpopulations through mutation or heritable phenotypic change [ 26 ] . The second mechanism is signaling-based. Lauffenburger and Cozens showed that inoculum-density effects in mammalian cell culture can emerge from autocrine growth-factor secretion, diffusion and receptor sensing [ 18 ] . Their theory predicts an inoculum density below which growth cannot be sustained. Above this threshold, proliferation increases with population size because denser cultures build up stronger extracellular signals, enhancing the growth stimulus received by each cell. Although both mechanisms are mechanistically informative, each is specific in a way that limits its generality. The evolutionary route attributes density-dependent growth to changes in population composition through mutation or heritable phenotypic variation. It therefore cannot, on its own, explain density-dependent growth observed in clonal populations over timescales where genetic evolution is not the primary driver [ 15 ] . The autocrine route, by contrast, explains density dependence through a particular signaling implementation, typically involving ligand secretion, receptor binding, and downstream pathways such as MAPK. While powerful, this description must be reformulated when the relevant ligand, receptor system, or intracellular signaling architecture changes. What is missing is a mesoscopic description that captures the common functional structure of these mechanisms without committing to a specific molecular realization. Although the two mechanisms operate through different biological processes, we propose that they share a common functional structure. At a mesoscopic level of description—above the molecular details of signaling pathways but below macroscopic population dynamics—the key effect of population size is to alter the relationship between the environmental signals available to a cell and the signals it actually senses. We treat this sensing discrepancy as a coarse-grained variable that abstracts away the specific molecular machinery responsible for it. Under this hypothesis, seemingly distinct mechanisms can be described within a common framework. Population size modifies the information available to individual cells, and cells respond by adjusting their phenotypic state to remain adapted to their environment. This perspective motivates modeling the cell as a Bayesian adaptive agent that updates its internal representation in response to sensory information. We do not suggest that cells literally perform Bayesian inference; rather, Bayesian updating provides a minimal coarse-grained description that captures the information-theoretic structure of sensing-driven adaptation [ 3 , 5 , 11 , 2 , 21 , 29 ] . Within this view, the cell is described by a coarse-grained internal phenotypic state that evolves on an intrinsic regulatory landscape. For a fixed genotype, this landscape encodes the cell’s regulatory tendency in the absence of environmental input, with a minimum at the maximally proliferative phenotype. This formulation is consistent with the biological default-state view that proliferation is a constitutive cellular tendency constrained by intracellular regulation rather than induced from zero [ 31 , 13 , 14 , 30 ] . Autocrine sensing then enters as a Bayesian inference-based update of phenotypic states: phenotypes whose sensed signal statistics better match the true extracellular signal are reinforced, while those that match poorly are depleted. The mismatch between sensed and true signal distributions therefore acts as a population-size-dependent bias on the intrinsic landscape, displacing the steady-state phenotype away from the proliferative optimum and reducing the average proliferation rate. Within this coarse-grained inference framework, a range of population-level growth behaviors arises from a common mechanism rather than being imposed as separate phenomenological assumptions: reduced survival at small population sizes, an intermediate proliferation optimum, a tissue-specific capacity, and superlinear scaling of growth at small population sizes. Our work complements rather than displaces evolutionary explanations of non-exponential tumor growth. Mutation may certainly contribute, but density-dependent proliferation can also arise from the way in which a clonal population statistically infers its microenvironment under constraints imposed by internal regulatory architecture The remainder of this paper is organized as follows. In Sec. 2, we develop the mathematical framework, formalizing phenotypic adaptation as a Bayesian update over an intrinsic regulatory landscape and deriving the stationary phenotype distribution in the