Agent-based Computational Model Research Articles

Scaling of agent-based models to evaluate transmission risks of infectious diseases

The scaling behaviour of agent-based computational models, to evaluate transmission risks of infectious diseases, is addressed. To this end we use an existing computational code, made available in the public domain by its author, to analyse the system dynamics from a general perspective. The goal being to obtain deeper insight into the system behaviour than can be obtained from considering raw data alone. The data analysis collapses the output data for infection numbers and leads to closed-form expressions for the results. It is found that two parameters are sufficient to summarize the system development and the scaling of the data. One of the parameters characterizes the overall system dynamics. It represents a scaling factor for time when expressed in iteration steps of the computational code. The other parameter identifies the instant when the system adopts its maximum infection rate. The data analysis methodology presented constitutes a means for a quantitative intercomparison of predictions for infection numbers, and infection dynamics, for data produced by different models and can enable a quantitative comparison to real-world data.

On the Epidemiological Evolution of Colistin-ResistantAcinetobacter Baumanniiin the City of Valencia: An Agent-Based Modelling Approach

Antibiotic resistance is one of the greatest public health threats today, mainly due to the non-rational use of antibiotics.Acinetobacter baumanniiis an example of a microorganism with high antibiotic resistance that has developed rapidly in recent years. Consequently, only a few lastresort antibiotics, such as colistin, are currently effective against it. In this work, we propose a random agent-based computational model to describe the evolution of colistin-resistantA. baumanniiin the population of Valencia (Spain) and to predict its impact both on the whole population and by age groups. The agent- based model uses a synthetic population of individuals with a vector of characteristics or state variables. These variables change over time based on a series of random events with certain conditional probabilities. The synthetic population statistical features and the probabilities have been found in demographic and hospital databases. One of these probabilities, the probability of infection by a resistant strain, has been modeled using random differential equations. The model takes into account antibiotic consumption as the primary driving force of variation and assumes non-rewersibility of resistance as the worst-case scenario. The agent-based model calibration and the selection of a real-world representative set of solutions have been carried out using the Partide Swarm Optimization evolutionary algorithm. This approach takes into account the inherent stochasticity of the model and the uncertainty of the data. Finally, projections of the incidence and absolute cases of colistin-resistantA. baumanniihave been performed. Our results suggest that, if the same consumption pattern continues, the ervolution of the colistin-resistant strain proportion will be exponential, exceeding 50% in 2025. Additionally, the results reveal that, despite the low incidence in Valencian hospitals, the impact on people over 60 years old will be more significant in terms of the number of cases. Based on these findings, it can be deduced that colistin will cense to be an effective antibiotic in the coming years, negatively impacting the human population, especially the most advanced age groups.

Cellular-automaton model for tumor growth dynamics: Virtualization of different scenarios

Mathematical Oncology has emerged as a research field that applies either continuous or discrete models to mathematically describe cancer-related phenomena. Such methods are usually expressed in terms of differential equations, however tumor composition involves specific cellular structure and can demonstrate probabilistic nature, often requiring tailor-made approaches. In this context, cell-based models allow monitoring independent single parameters, which might vary in both time and space. By relying on extant tumor growth models in the literature, this study introduces cellular-automata simulation strategies that admit heterogeneous cell population while capturing both single-cell and cluster-cell behaviors. In this agent-based computational model, tumor cells are limited to follow four possible courses of action, namely: proliferation, migration, apoptosis or quiescence. Despite the apparent simplicity of those actions, the model can represent different complex tumor features depending on parameter settings. This study virtualized five different scenarios, showcasing model capabilities of representing tumor dynamics including alternate dormancy periods, cell death instability and cluster formation. Implementation techniques are also explored together with prospective model expansion towards deterministic features. The proposed stochastic cellular automaton model is able to effectively simulate different scenarios regarding tumor growth effectively, figuring as an interesting tool for in silico modeling, with promising capabilities of expansion to support research in mathematical oncology, thus improving diagnosis tools and/or personalized treatment.

Agent-based modeling predicts RAC1 is critical for ovarian cancer metastasis.

Experimental and computational studies pinpoint rate-limiting step(s) in metastasis governed by Rac1. Using ovarian cancer cell and animal models, Rac1 expression was manipulated, and quantitative measurements of cell-cell and cell-substrate adhesion, cell invasion, mesothelial clearance, and peritoneal tumor growth discriminated the tumor behaviors most highly influenced by Rac1. The experimental data were used to parameterize an agent-based computational model simulating peritoneal niche colonization, intravasation, and hematogenous metastasis to distant organs. Increased ovarian cancer cell survival afforded by the more rapid adhesion and intravasation upon Rac1 overexpression is predicted to increase the numbers of and the rates at which tumor cells are disseminated to distant sites. Surprisingly, crowding of cancer cells along the blood vessel was found to decrease the numbers of cells reaching a distant niche irrespective of Rac1 overexpression or knockdown, suggesting that sites for tumor cell intravasation are rate limiting and become accessible if cells intravasate rapidly or are displaced due to diminished viability. Modeling predictions were confirmed through animal studies of Rac1-dependent metastasis to the lung. Collectively, the experimental and modeling approaches identify cell adhesion, rapid intravasation, and survival in the blood as parameters in the ovarian metastatic cascade that are most critically dependent on Rac1.

How does cross-sectional sampling bias our understanding of adolescent romantic relationships?: An agent-based simulation.

Adolescent romantic relationships are developmentally significant, but relatively brief and often disrupted by changes in context. Large individual differences and age-related change make sampling complex. Most adolescents have multiple romantic relationships. Which should we sample? To better understand the issues involved, this study used a simulation - an agent-based computational model - to generate model worlds, each following the relationships formed and dissolved over 5 years. Cross-sectional sample estimates of the number, duration, and type of relationships were compared to population parameters of all relationships formed within the 5 years. Computational models can provide useful insight into sampling bias because (1) the processes producing the results are explicit, (2) results can be replicated to reduce sample idiosyncrasies, and (3) sample statistics can be compared to known population parameters. 1000 iterations were run of an agent-based model following 1000 individuals interacting for 60 "months." The model included three types of individuals differing in relationship duration. Two sets of 1000 cross-sectional samples were drawn from the 60,000 cross-sectional "months." Sample statistics were compared to the population parameters. Cross-sectional samples systematically over-represented longer relationships. The ability to detect individual differences in the duration and number of partners varied with time. These results suggest that cross-sectional survey and observational studies may be time sensitive and systematically distort our understanding of adolescent romantic relationships by oversampling longer-term relationships. Results also illustrate how computational models can provide insight into complex phenomena.

Measuring the impact of nonpharmaceutical interventions on the SARS-CoV-2 pandemic at a city level: An agent-based computational modelling study of the City of Natal.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic hit almost all cities in Brazil in early 2020 and lasted for several months. Despite the effort of local state and municipal governments, an inhomogeneous nationwide response resulted in a death toll amongst the highest recorded globally. To evaluate the impact of the nonpharmaceutical governmental interventions applied by different cities-such as the closure of schools and businesses in general-in the evolution and epidemic spread of SARS-CoV-2, we constructed a full-sized agent-based epidemiological model adjusted to the singularities of particular cities. The model incorporates detailed demographic information, mobility networks segregated by economic segments, and restricting bills enacted during the pandemic period. As a case study, we analyzed the early response of the City of Natal-a midsized state capital-to the pandemic. Although our results indicate that the government response could be improved, the restrictive mobility acts saved many lives. The simulations show that a detailed analysis of alternative scenarios can inform policymakers about the most relevant measures for similar pandemic surges and help develop future response protocols.

An agent-based model of social care provision during the early stages of Covid-19

Social care is a frequent topic in UK policy debates, with widespread concern that the country will be unable to face the challenges posed by the increase in demand for social care. While this is a societal problem whose dynamics depends on long-term trends, such as the increase of human lifespans and the drop of birth-rates, a short-term crisis, such as a pandemic, can affect the need and supply of social care to a considerable, although temporary, extent. Building on previous modelling effort of social care provision, we present an agent-based computational model to investigate social care provision in the context of a pandemic (using as an example, the early stages of the Covid-19 pandemic), and related mitigation policies, on social care demand and supply, using a proof-of-concept agent-based model (ABM). We show how policy solutions aimed at controlling the pandemic may have substantial effects on the level of unmet social care need and propose that such models may help policymakers to compare alternative containment policies, taking into account their side effects on the social care provision process.

WITHDRAWN: Computational modeling of multiple myeloma growth and tumor aggregate formation

The mechanical signals received from the multiple myeloma microenvironment play a key role in bone marrow cancer development. However, the complexity of the cues received hinders the development of a novel and effective treatment. As such, agent-based computational models that consider coupled fluid and mechanical dynamics for every single cell may provide distinct perspectives and key information to adequately identify the specific tumor microenvironment. We have developed a novel 3D agent-based computational model that considers coupled fluid and particle dynamics to study multiple myeloma cell growth in response to the stimuli received from the cell microenvironment. Cell processes such as cell–cell interactions, cell maturation, and cell proliferation were considered by implementing user-defined functions in the commercial software Fluent. To calibrate and validate the developed model, we performed two experiments comparing the results of cell sedimentation velocity and cell proliferation with in vitro results. The results of the model were also compared with a finite element method model previously developed in-house. As the cells proliferated, the number of cell–cell and cell–extracellular matrix contacts increased ( ∼ 6.1% per hour), which reduced the cell maturation time ( ∼ 69.7%). Once the cells started to form aggregates, the proliferation process speeded up ( ∼ 9.9% per hour). Saturation of cell proliferation was observed after long-term simulation. Our results are qualitatively consistent with the in vitro results. In addition, they show that the cell proliferation rate increased as the number of cells considered increased. Furthermore, in a very dense aggregation, the lack of space prevented proliferation of internal cells. The model developed herein has significant advantages compared to the previous model in terms of computation, as it results in a significant reduction (84%–91%) in computational costs, thus allowing a more realistic simulation. • Agent-based model coupled with a continuous fluid model for Multiple Myeloma cell simulation. • Based on specific cell conditions, the model has been applied to study Multiple Myeloma cells’ growth and tumor aggregation formation. • Results have been validated with the literature, in-vitro experiments developed by the authors, and a previous in-house Finite Element model. • Cell maturation is stimulated by the increase in cell–cell and cell-ECM interactions, but it is inhibited by a lack of space in the inner parts of cell aggregations. • Computational costs have been reduced when compared to the previous Finite Element model.

Vocal learning, chorusing seal pups, and the evolution of rhythm

The presence of vocal learning and rhythm capacities in only a few mammals drives a cross-species hypothesis in evolutionary neuroscience: vocal learning and rhythm perception and synchronization may be causally related. Harbor seals (Phoca vitulina) are among these mammals, and their puppyhood is the most active vocal period in their lives. Building on both comparative psychology and animal behavior, this research integrates methods from human cognition with attention to species’ ecological sensory niches. We present data from seal pups in four different setups: recordings of semi-natural vocal interactions in (1) individual and (2) group settings, (3) playback experiments to elicit vocal responses, and (4) perceptual listening experiments to measure behavioral responses. We complement empirical data with agent-based computational modeling aimed at reverse-engineering the putative mechanisms underlying rhythmic vocal interactions. Our data suggest that seal pups have developed capacities to produce and perceive rhythmic patterns. Seals’ vocal exchanges show rhythmic interactivity and antisynchronous coordination. Evidence for antisynchrony, rather than synchrony, in seals suggests that the rhythm-vocal learning link across species is more complex than previously surmised.

CDC1 coordinate innate and adaptive responses in the omentum required for T cell priming and memory.

In the peritoneal cavity, the omentum contains fat-associated lymphoid clusters (FALCs) whose role in response to infection is poorly understood. After intraperitoneal immunization with Toxoplasma gondii, conventional type 1 dendritic cells (cDC1s) were critical to induce innate sources of IFN-γ and cellular changes in the FALCs. Unexpectedly, infected peritoneal macrophages that migrated into the FALCs primed CD8+ T cells. Although T cell priming was cDC1 independent, these DCs were required for maximal CD8+ T cell expansion. An agent-based computational model and experimental data highlighted that cDC1s affected the magnitude of the proliferative burst and promoted CD8+ T cell expression of nutrient uptake receptors and cell survival. Thus, although FALCs lack the organization of secondary lymphoid organs, cDC1s resident in this tissue coordinate innate responses to microbial challenge and provide secondary signals required for T cell expansion and memory formation.

Durability of Aligned Microtubules Dependent on Persistence Length Determines Phase Transition and Pattern Formation in Collective Motion.

Collective motion is a ubiquitous phenomenon in nature. The collective motion of cytoskeleton filaments results mainly from dynamic collisions and alignments; however, the detailed mechanism of pattern formation still needs to be clarified. In particular, the influence of persistence length, which is a measure of filament flexibility, on collective motion is still unclear and lacks experimental verifications although it is likely to directly affect the orientational flexibility of filaments. Here, we investigated the collective motion of microtubules with different persistence lengths using a microtubule-kinesin motility system. We showed that local interactions between microtubules significantly vary depending on their persistence length. We demonstrated that the bundling of microtubules is enhanced by more durable alignment, rather than by greater likelihood of alignment. An agent-based computational model confirmed that the rigidity-dependent durability of microtubule alignment dominates their collective behavior.

Optimizing global COVID-19 vaccine allocation: An agent-based computational model of 148 countries.

BackgroundBased on the principles of equity and effectiveness, the World Health Organization and COVAX formulate vaccine allocation as a mathematical optimization problem. This study aims to solve the optimization problem using agent-based simulations.MethodsWe built open-sourced agent-based models to simulate virus transition among a demographically representative sample of 198 million people in 148 countries using advanced computational services. All countries continuing their current vaccine progress is defined as the baseline scenario. Comparison scenarios include achieving minimum vaccination rates and allocating vaccines based on pandemic levels.FindingsThe simulations are fitted using the pandemic data from 148 countries from January 2020 to June 2021. Under the baseline scenario, the world will add 24.36 million cases and 468,945 deaths during the projection period of three months. Inoculating at least 10%, 20%, and 26% of populations in all countries requires 1.12, 3.31, and 5.00 million additional vaccine doses every day, respectively. Achieving these benchmarks reduces new cases by 0.56, 2.74, and 3.32 million, respectively. If allocated by the current global distribution, 5.00 million additional vaccine doses will only avert 1.45 million new cases. If those 5.00 million vaccines are allocated based on projected cases in each country, the averted cases will increase more than six-fold to 9.20 million. Similar differences between allocation methods are observed in averted deaths.ConclusionThe global distribution of COVID-19 vaccines can be optimized to achieve better outcomes in terms of both equity and effectiveness. Alternative vaccine allocation methods may avert several times more cases and deaths than the current global distribution. With reasonable requirements on additional vaccines, COVAX could adopt alternative allocation strategies that reduce cross-country inequity and save more lives.

Is it Feasible to Teach Agent-based Computational Modeling to Elementary and Middle School Students?

Many of the computational models in science, engineering, and social studies are agent-based models. However, despite their growing importance and potential attractiveness to students, agent-based models are not taught in elementary, middle or high school. In this paper, the author explores the feasibility of teaching agent-based models to elementary and middle school students, and the extent students can learn modeling using agent-based mathematical and computational models.

Burgeoning Research in Emergent Leadership

This symposium features emerging developments in emergent leadership research, reflecting some of the following questions: What role can computational modeling play in exploring different theoretical perspectives of emergent leadership? What do the patterns of leader emergence within team networks reveal that single occurrences of the phenomenon cannot? What influence do cultural norms have on the effectiveness of informal leadership status-seeking behaviors? Of what importance is the gender of those being evaluated for emergent leadership status, and what about the evaluator? More specifically, the studies included in this symposium focus on utilizing agent-based computational modeling as a way to compare social-cognitive and social-interactionist perspectives of leadership emergence (Acton, Foti, & Lord), how social network analysis can reveal patterns of emergent leadership in teams, and how maintaining multiple leader relationships impacts team member stress (Choudhury & Maupin), how high-ability team members' self-promotive behaviors are interpreted differently based on cultural norms (Liu & Hanna), and how the gender of those granting leadership status, as well as the gender composition of informal teams, is predictive of emergent leadership evaluations (Farro, Kukenberger, & D'Innocenzo). The discussant, Dr. William Gardner, will discuss themes across studies, highlight future research directions, and lead an interactive discussion between presenters and the audience. Who recognizes women's emergent leadership? Multilevel effects of gender on emergent leadership Presenter: Andrea Farro; Rowan U. Presenter: Michael Kukenberger; U. of New Hampshire Presenter: Lauren D'Innocenzo; Drexel U. Using agent-based modeling to test core components of the leadership emergence process Presenter: Bryan Acton; Durham U. Business School Presenter: Roseanne Foti; Virginia Tech Presenter: Robert G. Lord; Durham U. Demands of informal leadership in self-managing teams: Stress and well-being in team members Presenter: Anwesha Choudhury; Binghamton U.-State U. of New York Presenter: Cynthia Kay Maupin; U. of Mississippi One impression to rule them all? Culture's role interpreting promotive tactics by emergent leaders Presenter: Bo Liu; Jiangxi U. of Fiance and Economics Presenter: Andrew A. Hanna; U. of Nebraska, Lincoln

Imagining the School of the Future Through Computational Simulations: Scenarios’ Sustainability and Agency as Keywords

Computational simulations are fundamental tools not only for scientific research but also for education. They are frequently used as virtual laboratories to foster students’ understanding of the theoretical concepts that lie at the basis of the simulated systems. Recent research works in STEM education have started to explore the potential of simulations as future-oriented objects, to support students in the development of future scenarios for real-world situations. In this paper, we present a teaching-learning module targeted to upper high-school students on simulations of complex systems. The peculiarity of this course is that, guiding the students through the conceptual and epistemological analysis of some computational agent-based models, we were able to ground on these disciplinary bases the introduction of key concepts of the futures studies, like that of scenario. More specifically, in this paper we address an original future-oriented activity in which the students were required to choose an urgent problem of their interest, imagine possible and desirable scenarios based on a simulation and identify the sequence of actions to be undertaken to reach the preferable future. In presenting the results of the module’s implementation we focus on two groups of students who spontaneously decided to address a problem related to the current educational system. In particular, we discuss how the future-oriented activity based on simulations led the students (i) to imagine sustainable scenarios for the school of the future, in which a dynamical equilibrium between opposite tensions is achieved, without any of them being eliminated and (ii) to recognize themselves as agents of transformation in a public, professional, and personal dimension.

A competitive advantage through fast dead matter elimination in confined cellular aggregates

Competition of different species or cell types for limited space is relevant in a variety of biological processes such as biofilm development, tissue morphogenesis and tumor growth. Predicting the outcome for non-adversarial competition of such growing active matter is non-trivial, as it depends on how processes like growth, proliferation and the degradation of cellular matter are regulated in confinement; regulation that happens even in the absence of competition to achieve the dynamic steady state known as homeostasis. Here, we show that passive by-products of the processes maintaining homeostasis can significantly alter fitness. Even for purely pressure-regulated growth and exclusively mechanical interactions, this enables cell types with lower homeostatic pressure to outcompete those with higher homeostatic pressure. We reveal that interfaces play a critical role for this specific kind of competition: there, growing matter with a higher proportion of active cells can better exploit local growth opportunities that continuously arise as the active processes keep the system out of mechanical equilibrium. We elucidate this effect in a theoretical toy model and test it in an agent-based computational model that includes finite-time mechanical persistence of dead cells and thereby decouples the density of growing cells from the homeostatic pressure. Our results suggest that self-organization of cellular aggregates into active and passive matter can be decisive for competition outcomes and that optimizing the proportion of growing (active) cells can be as important to survival as sensitivity to mechanical cues.

Dynamic instability of dendrite tips generates the highly branched morphologies of sensory neurons.

The highly ramified arbors of neuronal dendrites provide the substrate for the high connectivity and computational power of the brain. Altered dendritic morphology is associated with neuronal diseases. Many molecules have been shown to play crucial roles in shaping and maintaining dendrite morphology. However, the underlying principles by which molecular interactions generate branched morphologies are not understood. To elucidate these principles, we visualized the growth of dendrites throughout larval development of Drosophila sensory neurons and found that the tips of dendrites undergo dynamic instability, transitioning rapidly and stochastically between growing, shrinking, and paused states. By incorporating these measured dynamics into an agent-based computational model, we showed that the complex and highly variable dendritic morphologies of these cells are a consequence of the stochastic dynamics of their dendrite tips. These principles may generalize to branching of other neuronal cell types, as well as to branching at the subcellular and tissue levels.

The Invisible Primary in an Agent-Based Model: Ideology, Strategy, and Competitive Dynamics

Historical accounts of American presidential nominating contests suggest that candidates jockey over ideology and policy in ways that shape the outcomes of these races. Yet this aspect of competition has been difficult to analyze with the formal and statistical methods that dominate this research agenda. To address this gap, this article presents a computational agent-based model (ABM) of candidates’ ideological maneuvering during the invisible primary. We extend the framework developed by Michael Laver to study dynamic party competition in Europe, but recast it for the different context and to enable model fit to be more rigorously determined. Our analysis of data from the 2012 Republican invisible primary suggests the importance of ideological jockeying in this contest. Moreover, its dynamics can be well-explained by a basic version of the ABM in which candidates select between three strategies (aggregator, hunter or sticker) and then maintain that strategy over time. The fit of this model, particularly in the short run, can be improved by introducing a “momentum effect” that allows the candidates’ standing in the race to rise or fall without any accompanying ideological change.

Supporting Three-dimensional Learning on Ecosystems Using an Agent-Based Computer Model

Supporting Three-dimensional Learning on Ecosystems Using an Agent-Based Computer Model

Abstract A040: Intra-epithelia cell dynamics shape evolutionary dynamics and selection of therapy resistant clones in lung cancer

Abstract Lung cancer is a leading cause of cancer-related deaths. Despite enhanced characterization of genomic complexity and intra-tumor heterogeneity that underpin therapy failure, the mechanistic and predictive factors shaping and steering evolutionary dynamics remain less clear. In this study, we combined experimental, computational, and tumor molecular data analysis approach to investigate factors and rules governing evolutionary dynamics in lung cancer. A lineage tracing study in a BL/6 KRASG12D TP53mutant lung adenocarcinoma cell line model uncovers unexpected variation in the long-term fate of clones with no selective advantage or disadvantage, with clones arising near the edge of tumor colonies being favored to expand. An agent-based computational model of cell proliferation and cell-cell physical interaction, incorporating the experimentally driven assumption of compression-induced non-proliferation, recapitulated the growth patterns of tumor colonies and unequal expansion of clones. By further implementing cell random motility, the model predicts that enhanced motility destabilizes cell-cell cohesion and causes spatial clone intermixing with altered size distributions. Corroborating the observations in silico, experimental manipulations that increase the migratory capabilities of cancer cells relieve density-driven growth arrest and lead to clone intermixing in vitro. Furthermore, lung cancer cell lines with more migratory properties exhibit less propensity for suppression of proliferation. To relate findings in our models to tumor data and to infer the degree of clone intermixing in patients, we implemented a mutational process in the computational model and extracted features of variant allele frequency that enable to distinguish model runs with higher or lower degree of clone intermixing. Inference in lung adenocarcinomas in the TRACERx (TRAcking Cancer Evolution through therapy (Rx)) Lung study using these features, cross-referenced with RNA sequencing data in the same tumors, reveals notable correlation between mechanisms driving cell migration in our experimental models and inter-subclone mixing in clinical samples. Finally, modelling the competition between sensitive and mutagenesis-derived resistant clones under a virtual cytotoxic therapy predicts an accelerated selection of therapy resistant clones under conditions with enhanced cell motility. In summary, our integrated study demonstrates that intra-epithelia cell dynamics dictates the fates of clones both under neutral evolution and under selective pressures imposed by therapy. Citation Format: Xiao Fu, Ajay Bhargava, Sasha Bailey, Dhruva Biswas, Carlos Martinez Ruiz, Sunil Kumar, Paul French, Nicholas McGranahan, Charles Swanton, Paul A. Bates, Erik Sahai. Intra-epithelia cell dynamics shape evolutionary dynamics and selection of therapy resistant clones in lung cancer [abstract]. In: Proceedings of the AACR Special Conference on the Evolutionary Dynamics in Carcinogenesis and Response to Therapy; 2022 Mar 14-17. Philadelphia (PA): AACR; Cancer Res 2022;82(10 Suppl):Abstract nr A040.