Organized by: Hana Dobrovolny (Texas Christian University, United States), Gilberto Gonzalez-Parra (New Mexico Tech, United States)
Note: this minisymposia has multiple sessions. The second session is MS02-IMMU.

• Guang Lin (Purdue University, United States)
  "Predicting the COVID-19 pandemic with uncertainties using data-driven models"
We have developed an integer-order COVID-19 epidemic model and a fractional-order COVID-19 epidemic model to reconstruct and forecast the transmission dynamics of COVID-19 in New York City. To quantify the uncertainties in the proposed data-driven epidemic model, we have investigated model sensitivity analysis, structural and practical identifiability analysis, model calibration, and uncertainty quantification. We have employed Bayesian model calibration and physics-informed machine learning algorithms to calibrate the model parameters. In the early stage of the outbreak in New York City, the reproduction number was around 4.3, which indicates this outbreak has high transmissibility. We observed that multi-pronged interventions, such as the stay-at-home order and social distancing, had positive effects on controlling the outbreak and slowing the virus's spread. In addition, we employed the proposed data-driven models to evaluate the effects of various strategies to deploy the Covid-19 vaccine to control the pandemic. We have also applied the formulation to infer the dynamics of COVID-19 in other cities/states, where the spread dynamic is different from New York City.
• Ana Vivas-Barber (Norfolk State University, United States)
  "Using Seasonality and Variable Incubation Periods to Study the Impact of Including Domestic Animals on the Dynamics of Malaria Transmission"
We investigate the impact of including variable mosquito population and added variable long and short incubation periods on the transmission dynamics of Malaria in Korea. The SEIS model is based on malaria infected mosquitoes which bite humans or animals. This model studies plasmodium vivax malaria and has variables for animal population and mosquito attraction to animals. The basic reproduction number of the ODE model with seasonal mosquito population (exponential) is presented and analyzed. The existing time-independent Malaria population ODE model was extended to time-dependent model with the difference explored. Also, using bi-modal Malaria incubation, changes to the infectious population when constant incubation period is extended to varied in the ODE model. Endemic equilibrium and stability analysis for the model was conducted with conditions on variables to insure solvability and DFE.
• Imelda Trejo Lorenzo (Los Alamos National Laboratory, United States)
  "A modified Susceptible-Infected-Recovered model for observed under-reported incidence data"
In this talk, I will present a mathematical model to quantify the fraction of unreported infected individuals during epidemic outbreaks. The model consists of three parts (1) a dynamical system base on the classical Susceptible-Infected-Recovered (SIR) epidemic model, (2) a stochastic model for the observed incidence and (3) a Bayesian approach to estimate the model parameters. We use the model to estimate the infection rate and fraction of under-reported individuals for the current Coronavirus 2019 outbreak in some American Countries. Our analysis reveals that consistently, about 50% of infected individuals were not observed in various South American outbreaks.
• Naveen Vaidya (San Diego State University, United States)
  "Modeling the risk of SARS-CoV-2 transmission from fomites"
The novel coronavirus disease (COVID-19) constitutes one of the most devastating pandemics of the 21st century. While direct person-to-person transmission of SARS-CoV-2, the etiological agent of COVID-19, appears to be the primary route of transmission, the contraction of SARS-CoV-2 from fomites in the environment is also considered a potential contributor to the disease transmission. In this talk, I will present a mathematical model to predict the probability of detecting SARS-CoV-2 in the environmental reservoirs during the COVID-19 outbreak in a community. Furthermore, we extend our model to predict the potential contribution of fomite transmission to the generation of new COVID-19 cases. We validate our model using experimental data with a large number of swab samples collected from commonly touched surfaces across San Diego County. Our model, which is capable of describing transmission dynamics of COVID-19 within San Diego county, allows us to compute the risk for an individual to encounter virus in the environment. The results indicate that the persistence of virus in some environmental surfaces can lead to a significant number of COVID-19 cases in the community.

Organized by: Hana Dobrovolny (Texas Christian University, United States), Gilberto Gonzalez-Parra (New Mexico Tech, United States)
Note: this minisymposia has multiple sessions. The second session is MS01-IMMU.

• Benito Chen-Charpentier (University of Texas at Arlington, United States)
  "Deterministic and stochastic modeling of plant virus propagation with delay"
Plant diseases caused by a virus are mostly transmitted by a vector that bites an infected plant and bites a susceptible one. There is a delay between the time a plant gets bitten by an infected vector and the time it is infected. In this paper we consider two simple models of plant virus propagation and study different ways in which delays can be incorporated including the addition of an exposed class for the plants. Simulations are done and comparisons with the results for the models without delays are presented.
• Kenichi Okamoto (University of St. Thomas, United States)
  "Opposing within-host and between-host selection pressures for virulence: Implications for disease surveillance"
For many infectious diseases, including SARS-Coronavirus-2 (SARS-CoV-2), disease surveillance followed by isolating, contact-tracing and quarantining infectious individuals is critical for controlling outbreaks. These interventions often begin by identifying symptomatic individuals. However, by actively removing pathogen strains likely to be symptomatic, such interventions may inadvertently select for strains less likely to result in symptomatic infections. Additionally, the pathogen’s fitness landscape is structured around a heterogeneous host pool. In particular, uneven surveillance efforts and distinct transmission risks across host classes can drastically alter selection pressures. Here we explore this interplay between evolution caused by disease control efforts, on the one hand, and host heterogeneity in the efficacy of public health interventions on the other, on whether less symptomatic, but widespread, pathogens evolving. Using an evolutionary epidemiology model parameterized for coronaviruses, we show that symptoms-driven disease control ultimately shifts the pathogen’s fitness landscape to select for asymptomatic strains. We find such outcomes result when isolation and quarantine efforts are intense, but insufficient for suppression. Moreover, when host removal depends on the prevalence of symptomatic infections, intense isolation efforts can select for the emergence and extensive spread of more asymptomatic strains. The severity of selection pressure on pathogens caused by these interventions likely lies somewhere between the extremes of no intervention and thoroughly successful eradication. Identifying the levels of public health responses that facilitate selection for asymptomatic pathogen strains is therefore critical for calibrating disease suppression and surveillance efforts and for sustainably managing emerging infectious diseases.
• Baylor Fain (Texas Christian University, United States)
  "Validation of a GPU-based ABM for rapid simulation of viral infections"
We developed a new ABM/PDEM hybrid model for simulating virus spreading in a monolayer of a million cells. In this work, aspects of the simulations, such as the time step, are checked to verify the model is producing accurate data. Physical characteristics of the viral spread, such as the growth rate, decay rate, peak amount of virus, and time of peak virus, are compared with real data ranges for Influenza virus. Values for the parameters: viral production rate, rate of infection, amount of time in the eclipse phase, and the amount of time in the infectious phase, are found for H1N1pdm09-WT from fitting the model to experimental data by minimizing the SSR (Sum of Square Residuals).
• Hayriye Gulbudak (University of Louisiana at Lafayette, United States)
  "A Delay Model for Persistent Viral Infections in Replicating Cells"
Persistently infecting viruses remain within infected cells for a prolonged period of time without killing the cells and can reproduce via budding virus particles or passing on to daughter cells after division. The ability for populations of infected cells to be long-lived and replicate viral progeny through cell division may be critical for virus survival in examples such as HIV latent reservoirs, tumor oncolytic virotherapy, and non-virulent phages in microbial hosts. We consider a model for persistent viral infection within a replicating cell population with time delay modelling the length of time in the eclipse stage prior to infected cell replicative form. We obtain reproduction numbers that provide criteria for the existence and stability of the equilibria of the system. Moreover, we characterize bifurcations in our model, including transcritical (backward and forward), saddle-node, homoclinic, and Hopf bifurcations, and provide evidence of a Bogdanov-Takens bifurcation. We investigate the possibility of long-term survival of the infection (represented by chronically infected cells and free virus) in the cell population by using the mathematical concept of robust uniform persistence. Using numerical continuation software with parameter values estimated from phage-microbe systems, we obtain two parameter bifurcation diagrams that divide parameter space into regions with different dynamical outcomes. We thus investigate how varying different parameters, including how the time spent in the eclipse phase, can influence whether the virus survives.

Organized by: Adriana Zanca (The University of Melbourne, Australia), Jennifer Flegg (The University of Melbourne, Australia), Helen Byrne (University of Oxford, UK)
Note: this minisymposia has multiple sessions. The second session is MS04-IMMU.

• Alison Betts (Applied BioMath, USA)
  "Modeling strategies for preclinical to clinical translation of T cell engager bispecific antibodies: using math to unravel counter intuitive dose responses"
T cell engager (TCE) bispecific antibodies are a promising therapeutic approach for the treatment of cancer. They have a complex mechanism of action, binding to CD3 on T cells and a tumor associated antigen on tumor cells to form a trimolecular complex (trimer), mimicking the normal immune synapse. Trimer formation stimulates the T cell and redirects cytotoxicity against the tumor cell. This results in some interesting mechanistic behaviors, including bell shaped concentration response relationships, which can result in non-intuitive dose response relationships. To understand these complex quantitative relationships, and to provide a tool for decision making from early discovery through to clinical trials, a translational quantitative systems pharmacology (QSP) model is proposed for TCE molecules.  The model predicts trimer formation between drug, T-cell and tumor cell, which can be linked to downstream pharmacodynamics, efficacy or toxicity. Two case studies are discussed; in the first the model is used to optimize design of a PSMA/CD3 TCE and in the second the model is used for preclinical to clinical translation of a Pcad/CD3 TCE to predict clinical efficacious dose.
• Allison Lewis (Lafayette College, USA)
  "Bayesian information-theoretic calibration of tumor models for informing effective scanning protocols"
With new advancements in technology, we can now collect data describing tumor growth using numerous metrics. For any tumor growth model, we observe large variability among individual patients’ parameter values, particularly those relating to treatment response; thus, exploiting the use of these various metrics for model calibration can be helpful to infer such patient-specific parameters both accurately and early. Since clinicians are limited to a sparse collection schedule, the determination of optimal times and metrics for which to collect data in order to best inform model calibration is essential. Here, we employ a Bayesian information-theoretic calibration protocol for experimental design in order to identify the optimal times at which to collect data for informing treatment parameters. Data collection times are chosen sequentially to maximize the reduction in parameter uncertainty with each added measurement, ensuring that a budget of n measurements results in maximum information gain about the model parameter values.
• Leili Shahriyari (University of Massachusetts Amherst, USA)
  "A data-driven mathematical model of colon cancer"
Every colon cancer has its own unique characteristics, and therefore may respond differently to identical treatments. Here, we introduce a data driven mathematical model for the interaction network of key components of immune microenvironment in colon cancer. We estimate the relative abundance of each immune cell from gene expression profiles of tumors, and group patients based on their immune patterns. We then compare the tumor sensitivity and progression in each of these groups of patients and observe differences in the patterns of tumor growth as well as response to FOLFIRI treatment.
• Min Song (University of Southern California, USA)
  "Quantitative analysis of endothelial sprouting mediated by FGF- and VEGF-induced MAPK and PI3K/Akt pathways"
The essential role of blood vessels in delivering nutrients makes angiogenesis important in wound healing and tumor growth. Targeting angiogenesis is a prominent strategy in tissue engineering and cancer treatment. However, not all approaches to regulate angiogenesis lead to successful outcomes. There is a limited understanding of how pro-angiogenic factors such as VEGF and FGF combine together to stimulate angiogenesis. We aim to quantitatively characterize the crosstalk between VEGF- and FGF-mediated angiogenic signaling and endothelial sprouting, to gain mechanistic insights and identify novel therapeutic strategies. We constructed a hybrid agent-based model that characterizes endothelial sprouting driven by FGF and VEGF-mediated MAPK and PI3K/Akt signaling. The experimentally fitted and validated model predicts that FGF induces stronger angiogenic responses in the long-term compared to VEGF stimulation. Also, FGF plays a dominant role in the combination effects in endothelial sprouting. Moreover, the model suggests that ERK and Akt pathways and cellular responses contribute differently to the sprouting process. Furthermore, the model predicts that the strategies to modulate endothelial sprouting are context dependent. Thus, our model can identify potential effective pro- and anti-angiogenic targets under different conditions and study their efficacy. The model provides mechanistic insight into VEGF and FGF interactions in sprouting angiogenesis.

Organized by: Adriana Zanca (The University of Melbourne, Australia), Jennifer Flegg (The University of Melbourne, Australia), Helen Byrne (University of Oxford, UK)
Note: this minisymposia has multiple sessions. The second session is MS03-IMMU.

• Wafaa Mansoor (Murdoch University, Australia)
  "Modelling hydrogen clearance in the retina"
Two simple mathematical models of advection and diffusion of hydrogen within the retina are discussed to assist in interpretation of the ’hydrogen clearance technique’ that is used to estimate blood flow. The first model assumes the retina consists of three, well-mixed layers with different thickness, two-dimensional model consisting of three regions that represent the layers in the retina. Diffusion between the layers and leakage through the outer edges are considered. Solutions to the governing equations are obtained by employing Fourier series and finite difference methods for the two models, respectively. The effect of important parameters on the hydrogen concentration is examined and discussed. The results contribute to understanding the dispersal of hydrogen in the retina and in particular the effect of flow in the vascular retina. It is shown that the predominant features of the process are captured by the simpler model.
• Vijayalakshmi Srinivasan (Auckland Bioengineering Institute, New Zealand)
  "3D analysis of Human placental cotyledon: a step ahead to understand feto-placental vasculature"
The human placenta has extensive branching villus structure, which contains a branching network of fetal blood vessels that are essential for efficient exchange of nutrients from mother to fetus. Reduced vascular density and branching have been linked to functional placental impairments, such as fetal growth restriction (FGR) where the baby’s growth rate becomes dangerously reduced. Currently, we lack clear understanding of origins of FGR for early diagnosis and treatment. Computational models that mimic the structure and function of the feto-placental vasculature have proved useful in predicting the consequences of perturbations to these structures in FGR. However, they have been limited in their anatomical fidelity at the meso-scale (the primary site of resistance), due to challenges in imaging the placenta. Here, we present our approach to simulating feto-placental vascular function in the placenta as a whole, which aims to accurately incorporate structural detail regarding branching properties of the complex vascular tree. We then present new data regarding the complexity of the feto-placental vasculature at the cotyledon (functional unit) scale, and show how mathematical models representing the cotyledon as a branching network of vessels can be used to interrogate function across spatial scales relevant to the key sites of feto-placental vascular resistance.
• Yuhuang Wu (Kirby Institute, Australia)
  "Predicting the composition of the HIV / SIV Reservoir and Rebounding Virus"
Human Immunodeficiency Virus (HIV) attacks human immune cells and new free virus is produced via infected cells. Even with a successful treatment of HIV, the population of infected cells does not go extinct. Instead, a number of infected cells stay in an inactive state and once treatment is stopped, reactivation of infected cells may produce virus again. To date, it is still unclear when and how these inactive infected cells (reservoir) are formed. In this work, we try to distinguish whether different virus produced throughout the course of infection contributes equally to the formation of reservoir, or virus strains produced over a certain time period are more important to reservoir formation. Furthermore, we explore how the composition of replicating virus relates to the composition of the reservoir. Additionally, we look at if the reservoir composition determines the production of virus when the treatment is stopped. In this talk, we use both mathematical modelling and statistical analysis, applied to experimental data from an animal study, to show the relationship between the early viral dynamics and the reservoir composition as well as the recrudescent virus. We find dominant viral strains present prior to treatment are more likely to reactivate after cessation of treatment.
• Claire Miller (University of Amsterdam, Netherlands)
  "In silico clinical trials for acute ischemic stroke"
The concept of in silico trials is gaining increasing attention in medical research. The end goal of these trials is to refine, reduce the cost of, and partially replace in vivo animal studies and human clinical trials. In our project, INSIST, we are developing in silico trials for acute ischemic stroke (AIS): the occlusion of an artery in the brain. The current standard of care for AIS is thrombolysis (drug) and/or thrombectomy (surgical) intervention. Modelling AIS requires the modelling of the stroke onset, treatment, and resulting injury. This is done by linking models for blood flow through the arteries, blood perfusion in the brain, the two treatment approaches, and tissue injury. It is also necessary to be able to produce large numbers of patients to run these models on; provide trial outcomes that are clinically relevant; and a trial framework that can be practically compared to current traditional clinical trials. In this talk I will discuss the setup of the INSIST in silico trials, how we connect the different models to predict treatment and patient outcome, and the methods we have used to generate populations of virtual patients using clinical data. Additionally I will discuss how the approaches used facilitate the translation of the trial outcomes to a clinical setting.

Organized by: Roya Zandi (University of CA, Riverside, USA), Amber Smith (University of Tennessee, USA), Reidun Twarock (University of York, UK)

• Carolyn Teschke (Department of Molecular and Cell Biology, University of Connecticut, USA)
  "Using a scaffold to build a viral capsid"
Many dsDNA viruses, including the herpesviruses and tailed bacteriophages, build a precursor capsid called a procapsid into which the dsDNA is subsequently packaged. These viruses require an internal scaffolding protein to assemble coat proteins into procapsids of the proper size and shape. How a scaffolding protein affects the conformation of a coat protein so that it is competent for assembly is not understood. We have used single molecule fluorescence experiments to demonstrate a surprisingly high affinity interaction between bacteriophage P22’s monomeric scaffolding protein and coat protein. This interaction shifts coat protein into a conformation consistent with the procapsid configuration of the protein. Thus, scaffolding protein directly activates coat protein for assembly.
• Siyu Li (Northwestern University, China)
  "The physical mechanism of virus self-assembly"
Understanding the basic mechanism of virus self-assembly is fundamental in deciphering the formation and evolution of viruses and exploring their applications to drug delivery, gene therapy and vaccination. While considerable progress has been achieved in determining the virus structures, kinetic pathways by which hundreds or thousands of proteins assemble to form structures with icosahedral order (IO) is still elusive. To decipher the assembly pathway, we developed a computational model to simulate the virus growth. We proposed two mechanisms of small virus assembly: en-mass and nucleation-growth, and studied the role of elasticity and genome in the disorder-order transition process. Moreover, we study the growth of large viruses and discover the universal role of scaffolding proteins in the formation of viral capsids. Using continuum elasticity theory, we show that a nonspecific template not only selects the radius of the capsid, but also leads to the error-free assembly of protein subunits into capsids with universal IO. The mechanism we study will help us deeply understand the correlation between protein building blocks and virus macrostructures, and guide the experiments to explore the possibility of antiviral drugs that inhibit the virus self-assembly.
• Trevor Douglas (Department of Chemistry, Indiana University, Bloomington IN 47405, USA)
  "Directed Assembly of Virus-Based Nanoreactors Across Multiple Lengthscales"
The virus like particles (VLP) derived from the bacteriophage P22 provide an opportunity for constructing catalytically functional nano-materials by directed encapsulation of enzyme cargos into the interior volume of the capsid. Directed enzyme encapsulation is achieved by genetically fusing the enzyme of interest to a truncated version of the scaffolding protein, which directs capsid assembly and is encapsulated within the capsid. This approach affords the packing of the desired enzymes within the roughly 60 nm diameter P22 capsid at very high packing density. The self-assembly of these nanoreactors is dependent on the multivalent nature of the cargo and this can be used to control the density of encapsulated cargo. We have explored the molecular level porosity of capsid and determined the range of substrates that can access the encapsulated enzyme and the dependence of this gating on molecular size and charge. Using these P22 nanoreactors as individual building blocks we can extend their utility towards the fabrication of hierarchically complex systems by further manipulation of their exterior surfaces. Superlattice materials, with long-range order, can be assembled through the directed hierarchical assembly of individual P22 particles, mediated by interparticle electrostatic interactions or through interactions of surface bound decoration proteins. In this way we can create ordered 3D materials that exhibit complex coupled behavior through communication between individual P22 nanoreactors.
• Giuliana Indelicato (Department of Mathematics, The University of York, UK)
  "The role of surface stress in non-quasi-equivalent viral capsids"
We focus here on viruses in the PRD1-adenovirus lineage which do not always conform to the Caspar and Klug classification. Instead of being built from one type of capsid protein (CP), they either code for multiple distinct structural proteins that break the local symmetry of the capsomers in specific positions, or exhibit auxiliary proteins that stabilize the capsid shell. We investigate the hypothesis that this occurs as a response to mechanical stress. We construct a coarse-grained model of a viral capsid, derived from the experimentally determined atomistic positions of the CPs. For T = 28 viruses in this lineage, which have capsids formed from two distinct structural proteins, we show that the tangential shear stress in the viral capsid concentrates at the sites of local symmetry breaking. In the T = 21, 25 and 27 capsids, we show that stabilizing proteins decrease the tangential stress. These results suggest that mechanical properties can act as selective pressures on the evolution of capsid components, counterbalancing the coding cost imposed by the need for such additional protein components.

Organized by: Jennifer Mueller (Colorado State University, United States)

• Bradford Smith (Department of Bioengineering, University of Colorado Denver | Anschutz Medical Campus, United States)
  "Ventilator waveform analysis to diagnose and prevent ventilator-induced lung injury"
Acute respiratory distress syndrome (ARDS) is caused by diverse factors including sepsis, trauma, and COVID-19. The derangements of lung function associated with ARDS necessitate mechanical ventilation to sustain life. However, the mechanical ventilator can also cause additional ventilator-induced lung injury that leads to worse ARDS outcomes. Adjusting the mechanical ventilator to minimize VILI is a challenging task because the injurious forces are functions of the applied ventilation and the mechanical properties of the lung which, in turn, depend on injury severity and type. As such, the optimal ventilation settings for each patient are likely different, change with time, and are not readily discernable from clinical data. To address this challenge, our long-term goal is to develop a system to numerically identify and apply the optimally lung-protective ventilation for any particular patient. The first step is to develop and validate simulations that can accurately predict the response of the injured lung to changes in ventilator settings. We have developed a compartment model of the respiratory system that accounts for nonlinear tissue elastance, lung resistance, and the nonlinear dynamics of alveolar recruitment. The model parameters are identified by fitting to pressures and volumes measured in mechanically ventilated mice (the training data). The model predictions are compared to evaluation data collected in the same animal to show that this approach provides accurate predictions of the response of the injured lung to ventilator adjustments. The model outputs also provide an accurate assessment of lung injury severity when compared to gold-standard lung function assessments performed using flexiVent research ventilators.
• Emily Heavner (Colorado State University, United States)
  "Estimation of airway resistance throughout the bronchial tree from mechanical ventilation output data"
We introduce a multi-compartment lung model based on resistance-capacitor circuits using an analogy between electric circuits and the human lungs. Multiple literature sources reveal a wide range of clinically used values for airway resistance, motivating an investigation to determine the role of airway resistance in the alveolar tree. The inverse problem of computing the vector of airway resistance values in the alveolar tree is studied using a linear least squares optimization approach. We compare the outputs of the model to real-world parameters collected from mechanical ventilation data of COVID-19 positive and negative patients.
• Bela Suki (Dept. Biomedical Engineering, Boston University, United States)
  "Inflation instability in the lung: An analytical model of a thick-walled alveolus with wavy fibers under large deformations"
Inflation of hollow elastic structures can become unstable and exhibit a runaway phenomenon if the tension in their walls does not rise rapidly enough with increasing volume. Biological systems avoid such inflation instability for reasons that remain poorly understood. This is best exemplified by the lung, which inflates over its functional volume range without instability. The goal of this study was to determine how the constituents of lung parenchyma determine tissue stresses that protect alveoli from instability-related over-distension during inflation. We present an analytical model of a thick-walled alveolus composed of wavy elastic fibers, and investigate its pressure-volume behavior under large deformations. Using second harmonic generation imaging, we found that collagen waviness follows a beta distribution. Using this distribution to describe human pressure-volume curves, we estimated collagen and elastin effective stiffnesses to be 1247 and 18.3 kPa, respectively. Furthermore, we demonstrate that linearly elastic but wavy collagen fibers are sufficient to achieve inflation stability within the physiological pressure range if the alveolar thickness-to-radius ratio > 0.05. Our model thus identifies the constraints on alveolar geometry and collagen waviness required for inflation stability and provides a multiscale link between alveolar pressure and stresses on fibers in healthy and diseased lungs.
• Vitor Mori (University of Vermont, United States)
  "Modelling the progression of Ventilation-Induced Lung Injury in Mice"
Mechanical ventilation is a crucial tool in the management of acute respiratory distress syndrome, yet it may itself also further damage the lung in a phenomenon known as ventilator-induced lung injury (VILI). We have previously shown in mice that volutrauma and atelectrauma act synergistically to cause VILI. We have also postulated that this synergy arises because of a rich-get-richer mechanism in which repetitive lung recruitment generates initial small holes in the blood-gas barrier which are then expanded by over-distension in a manner that favors large holes over small ones. In order to understand the causal link between this process and the derangements in lung mechanics associated with VILI, we developed a mathematical model that incorporates both atelectrauma and volutrauma to predict how the propensity of the lung to derecruit depends on the accumulation of plasma-derived fluid and proteins in the airspaces. We found that the model accurately predicts derecruitment in mice with experimentally induced VILI.

Organized by: Thomas Hillen (University of Alberta, Canada), Carlos Contreras (University of Alberta, Canada)
Note: this minisymposia has multiple sessions. The second session is MS10-IMMU.

• Morgan Craig (Sainte-Justine University Hospital Research Centre/Université de Montréal, Canada)
  " The impact of viral variants on immunopathology in COVID-19"
As SARS-CoV-2 continues its spread, the emergence of new variants has attracted increased attention, particularly as vaccination efforts ramped up. Throughout the pandemic, there has been a considerable effort to understand the genomic evolution of the virus. A quantitative picture of the evolution of SARS-CoV-2 in response to within-host pressures and their influence on the immunological response to infection is a crucial component to understanding and predicting COVID-19 outcomes. We have previously developed a mechanistic mathematical model of the immunological response to SARS-CoV-2 infection. Leveraging this framework, here we studied how viral variants influence immunopathology in COVID-19. Merging within-host SARS-CoV-2 evolutionary data and our cohort of realistic virtual patients, we predicted the combined effects of spike proteins and interferon-evading mutations on COVID-19 severity. Our results suggest that an individual’s immune response and their potential propensity for severe COVID-19 are the key factors distinguishing COVID-19 disease courses and outcomes.
• Ashlee N. Ford Versypt (University at Buffalo, The State University of New York, USA)
  "Multiscale Simulation of Lung Fibrosis Induced by SARS-CoV-2 Infection and Acute Respiratory Distress Syndrome"
The 2019 novel coronavirus, SARS-CoV-2, is a pathogen of critical significance to international public health. Knowledge about immune system-virus-tissue interactions and how these can result in low-level infections in some cases and acute respiratory distress syndrome (ARDS) and other tissue damage in others is limited. We are developing an open-source, multi-scale tissue simulator that can be used to investigate mechanisms of intracellular viral replication, infection of epithelial cells, host immune response, and tissue damage. Our model can simulate fibroblast-mediated collagen deposition to account for the fibrosis at the damaged site in response to immune-response-induced tissue injury. The severity of infection and collagen deposition depends on the anti-inflammatory cytokine secretion rate, multiplicity of infection, and contact time for a CD8+ T cell to kill an infected cell. Additionally, the change in the ACE2 receptor concentration from the multiscale model has been used in a separate model of renin-angiotensin system to predict the change in ANGII, which is a biomarker for hypertension, pro-inflammation, and pro-fibrosis.
• Paul Macklin (Indiana University, USA)
  "Community-driven multiscale model of SARS-CoV-2 dynamics and immune response"
The 2019 novel coronavirus, SARS-CoV-2, is a pathogen of critical significance to international public health. Knowledge of the interplay between molecular-scale virus-receptor interactions, single-cell viral replication, in-tracellular-scale viral transport, and emergent tissue-scale viral propagation is limited. Moreover, little is known about immune system-virus-tissue interactions and how these can result in low-level (asymptomatic) infections in some cases and acute respiratory distress syndrome (ARDS) in others, particularly with respect to presentation in different age groups or pre-existing inflammatory risk factors. Given the nonlinear interactions within and among each of these processes, multiscale simulation models can shed light on the emergent dynamics that lead to divergent outcomes, identify actionable “choke points” for pharmacologic interventions, screen potential therapies, and identify potential biomarkers that differentiate patient outcomes. Given the complexity of the problem and the acute need for an actionable model to guide therapy discovery and optimization, we introduce and iteratively refine a prototype of a multiscale model of SARS-CoV-2 dynamics in lung tissue. The first prototype model was built and shared internationally as open source code and an online interactive model in under 12 hours, and community domain expertise is driving regular refinements. In a sustained community effort, this consortium is integrating data and expertise across virology, immunology, mathematical biology, quantitative systems physiology, cloud and high performance computing, and other domains to accelerate our response to this critical threat to international health. More broadly, this effort is creating a reusable, modular framework for studying viral replication and immune response in tissues, which can also potentially be adapted to related problems in immunology and immunotherapy.
• Adrianne Jenner (Queensland University of Technology, Australia)
  "Virtual patient cohort reveals immune mechanism driving COVID-19 disease outcomes"
Manifestations of SARS-CoV-2 infection are heterogeneous, and a large proportion of people experience asymptomatic or mild infections that do not require hospitalization. In severe cases, patients develop coronavirus disease (COVID-19), which is frequently accompanied by a myriad of inflammatory indicators and hospitalization. To understand the diversity of immune responses to SARS-CoV-2 and distinguish features that predispose individuals to severe COVID-19, we developed a mathematical model (system of delay differential equations) and from that interpolated a virtual patient cohort. Our results indicate that virtual patients with low production rates of IFN subsequently experienced highly inflammatory disease phenotypes, compared to those with early and robust IFN responses. In these in silico patients, the concentration of interleukin-6 (IL-6) was also a major predictor of CD8+ T cell depletion (a known marker of disease severity in hospitalised patients). Our analyses predicted that individuals with severe COVID-19 also have accelerated monocyte-to-macrophage differentiation that was mediated by increased IL-6 and reduced type I IFN signalling. Together, these findings identify biomarkers driving the development of severe COVID-19 and support early interventions aimed at reducing inflammation.

Organized by: Thomas Hillen (University of Alberta, Canada), Carlos Contreras (University of Alberta, Canada)
Note: this minisymposia has multiple sessions. The second session is MS09-IMMU.

• Suzan Farhang Sardroodi (York University, Canada)
  "Analysis of host immunological response of adenovirus-based Covid-19 vaccines"
The coronavirus disease 2019 (COVID-19) caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) can be mitigated through safe and effective administration of vaccines. In this work, we provide a mathematical framework to investigate the mechanism of vaccine-induced cellular and humoral adaptive immunity. The model uses a system of simple ordinary differential equations to analyze the safety and efficacy of the vaccine. We confront our model to various vaccine doses in an attempt to understand different immunological profiles. An optimum solution is to compute a vaccination strategy of smaller dosage and longer delay that allows the highest efficacy while allowing supply of the vaccine to catch up with the demand. Model parameters are compared to clinical trial data on adenovirus-vectored vaccines against COVID-19 but could be adapted with different vaccine types such as mRNA, protein subunit, or multi-epitope vaccines.
• Dominik Wodarz (UC Irvine, USA)
  "The impact of viral variants on the immune response during COVID-19"
N/A
• Carlos Contreras (University of Alberta, Canada)
  "Personalized Virus Load Curves of SARS-CoV-2 Infection"
We introduce an explicit function that describes virus-load curves on a patient-specific level. This function is based on simple and intuitive model parameters. It allows virus load analysis without solving a full virus load dynamic model. We validate our model on data from influenza A as well as SARS-CoV-2 infection data for Macaque monkeys and humans. Further, we compare the virus load function to an established target model of virus dynamics, which shows an excellent fit. Our virus-load function offers a new way to analyse patient virus load data, and it can be used as input to higher level models for the physiological effects of a virus infection, for models of tissue damage, and to estimate patient risks.
• Jane Heffernan (York University, Canada)
  " A multi-scale model for SARS-CoV-2 infection"
N/A

Organized by: Alexey Tokarev (Рeoples’ Friendship University of Russia, Russia)

• Vitaly Volpert (CNRS, University Lyon, France)
  "Introduction to the pathophysiology of the coronavirus disease"
A short overview of the current knowledge on the disease progression and its possible complications will be presented.
• Anass Bouchnita (Department of Integrative Biology, University of Texas at Austin, USA)
  "Multiscale modelling of SARS-CoV-2 infection to study the role of innate and adaptive immune responses in healthy and immunocompromised individuals"
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection causes mild to severe outcomes depending on the balance of host immune response. The interaction between SARS-CoV-2 and the immune response is complex because it involves processes that span across several scales of biological hierarchy such as cells, tissues, organs, and the host. In this talk, we present a multiscale model that describes the interaction between SARS-CoV-2 and the immune response. In this model, dendritic cells are considered as individual objects that move within a section of the epithelial tissue and can be used by the virus to replicate and spread. They also secrete type I IFN which downregulates the production of the virus. At the same time, the model simulates the production of antigen-specific by lymph nodes as well as their interaction with infected cells and virions in the infection site. After model validation, we show that a moderately weak type I IFN could elicit a solid adaptive response that accelerates the virus's clearance. Numerical simulations suggest that the deficiency of naïve lymphocytes in immunocompromised individuals increases the persistence of the virus and exacerbates the disease's outcome.
• Bogdan Kazmierczak (Institute of Fundamental Technological Research, Polish Academy of Sciences, Poland)
  "Infection spreading in cell culture as a reaction-diffusion wave"
We formulate a reaction-diffusion system of equations modeling the progression of viral infection, e.g. of SARS-Cov viruses. Analytical and numerical results obtained in the framework of the model are in agreement with the 'in vitro' experimental findings.
• Alexey Tokarev (S.M. Nikolskii Mathematical Institute, Рeoples’ Friendship University of Russia (RUDN University), Russia)
  "Nonlinear dynamics in the homogeneous model of immune responses to SARS-CoV-2 virus"
Antiviral immune response is a highly nonlinear process governed by the cooperative behavior of variegated constituents of immune system. Depending on nature of virus, initial viral load, and patient peculiarities, infection can pass diversely and result from recovery to death. In the current pandemic of COVID-19 infection, in the part of patients the disease is complicated by abnormal inflammation response (hypercytokinemia, cytokine storm). We study the immune response to the SARS-CoV-2 virus by constructing the series of ODE-based mathematical models of different phases of this infection: (1) innate immune response, (2) innate plus adaptive immune response, (3) inflammation response. The innate immune response model shows the bistability and threshold properties, as well as possible oscillatory regime. The higher the initial viral load, the shorter is the incubation period and the higher is the maximal transient virus concentration. Depending on the effectiveness of antibodies production, the adaptive immune response can either fully eliminate the virus, or substantially postpone virus concentration burst with following higher virus concentration comparing to the case of innate response only. Inflammation response model also shows bistability and oscillatory behavior. We compare prediction of these models with clinical and epidemiological data. Finally, we study the duration of vaccine protection against the SARS-CoV-2 virus. This work was supported by the Ministry of Science and Higher Education of Russian Federation: agreement no. 075-03-2020-223/3 (FSSF-2020-0018).

Organized by: Stanca Ciupe (Virginia Tech, United States), Jessica Conway (Penn State University, USA), Amber Smith (University of Tennessee Health Science Center, USA), Jonathan Forde (Hobart and William Smith Colleges, USA)
Note: this minisymposia has multiple sessions. The second session is MS14-IMMU.

• James Faeder (University of Pittsburg, USA)
  "Multiscale Modeling of Viral Replication and Interferon-mediated Immune Responses"
Most intrahost models of viral infections track virus are built on ordinary differential equations that track viral and cell population but that simplify processes at the intracellular level. While these models have yielded key insights into the factors that affect viral load kinetics and have identified how factor such as timing and mechanism can determine treatment efficacy, there are several questions that require more detailed modeling of interactions at the molecular level. In particular, viral replication products and host signaling pathways interact in numerous ways that determine both the quantitative and qualitative outcomes of infection. For example, type I interferon (IFN) responses elicited by virus infection of cells in lymphoid tissues near the sites of infection not only mediate resistance of the infected cells to viral replication, but also may provide systemic resistance. In particular, with encephalitic alphaviruses, the antiviral state is stimulated in the brain early after peripheral infection. It is important to understand the characteristics and cell types involved in this early interferon stimulation as they may be protective from fatal disease. We have developed an experimental model in which the encephalitic alphavirus, eastern equine encephalitis virus (EEEV), infects various types of immune cells in an in vitro culture system. Using this system we are able to measure the kinetics of various steps in viral replication and host cell response, including induction of Type I IFNs and induction of IFN-regulated genes. We will use data from this experimental model to build and calibrate a computational model that will predict cell type specific IFN responses to viral infection and the potentially distal effects of this induction on mitigating viral infections. We will used this integrated experimental and model-based approach to identify key control mechanisms in viral and host dynamics that could be utilized for design of therapies to mitigate the effects of viral infection.
• Hana Dobrovolny (Texas Christian University, USA)
  "An ODE model of syncytia formation during viral infections"
Several viral infections are known to form syncytia, which are multinuclear cells created by cells that have fused together. Little is known, however, about how the syncytia alter viral dynamics. We use an ODE model to study how different assumptions about the viral production of syncytia and lifespan of syncytia change the resulting infection time course. We find that the effect of syncytia on viral titer is only apparent when the basic reproduction number for infection via syncytia formation is similar to the reproduction number for cell free viral transmission. When syncytia fusion rate is high, we find the presence of syncytia can lead to long-lasting infections if viral production is suppressed in syncytia.
• Daniel Reeves (Fred Hutchinson Cancer Research Center, USA)
  "Merging viral dynamics and phylogenetics reveals host-mediated selection may be sufficient, but not necessary, to explain within-host HIV evolution"
Modern HIV research depends crucially on both viral sequencing and population size measurements. To directly link mechanistic biological processes and evolutionary dynamics during HIV infection, we developed multiple within-host phylodynamic (wi-phy) models of HIV primary infection for comparative validation against viral load and evolutionary dynamics data. The most parsimonious and accurate model required no explicit immune selection, suggesting that the host adaptive immune system reduces viral load, but does not drive observed viral evolution. Rather, genetic drift primarily dictates fitness changes. These results hold during early infection. Moreover, during chronic infection — a setting in which adaptive immune selection has been observed -- viral fitness distributions are not largely different from in vitro distributions that emerge without adaptive immunity. Simulations highlight how phylogenetic inference must consider complex viral and immune-cell population dynamics to gain accurate mechanistic insights.
• Jessica Conway (Penn State University, USA)
  "Unified model of short- and long-term HIV viral rebound"
Antiretroviral therapy (ART) effectively controls HIV infection, suppressing HIV viral loads. Typically suspension of therapy is rapidly followed by rebound of viral loads to high, pre-therapy levels. Indeed, a recent study showed that approximately 90% of treatment interruption study participants show viral rebound within at most a few months of therapy suspension, but the remaining 10%, showed viral rebound some months, years, or maybe permanently, after ART suspension. Design of therapeutic interventions to expand this latter group are underway. However, an understanding of the heterogeneity in rebound dynamics, crucial in design of clinical trials to test these interventions, is lacking. We will discuss our branching process model to gain insight into these post-treatment dynamics. Specifically we provide theory that explains both short- and long-term viral rebounds, and post-treatment control, via a branching process model with time-inhomgeneous rates, validated with data from Li et al. (2016). We will discuss the associated biological interpretation and implications. Finally we will provide an example of how our modeling can be used to inform HIV treatment suspension study design.

Organized by: Stanca Ciupe (Virginia Tech, United States), Jessica Conway (Penn State University, USA), Amber Smith (University of Tennessee Health Science Center, USA), Jonathan Forde (Hobart and William Smith Colleges, USA)
Note: this minisymposia has multiple sessions. The second session is MS13-IMMU.

• Ivan Ramirez-Zuniga (University of Tennessee Health Science Center, USA)
  "A data-driven mathematical model of the role of energy in sepsis"
Mounting an adequate acute immune response against a pathogenic infection is energetically expensive. In an ideal scenario, this response may eradicate the infection but, in some cases, an imbalanced response may lead to sepsis. In this talk I will present a mathematical model that captures the dynamics of an immune response and its energy requirements to fight an infection. We calibrate our model with available animal data and identified key parameters for distinguishing between surviving and non-surviving subjects. On our analysis, we found that energy-related processes play a fundamental role in determining these outcomes. Moreover, we explore factors that modulate the inflammatory response across baseline and altered glucose conditions.
• Sarah Minucci (Virginia Commonwealth University, USA)
  "Mathematical modeling of ventilator-induced lung inflammation"
Despite the benefits of mechanical ventilators, prolonged or misuse of ventilation may lead to ventilation-associated/ventilation-induced lung injury (VILI). Lung insults, such as respiratory infections and lung injuries, can damage the pulmonary epithelium, with the most severe cases needing mechanical ventilation for effective breathing and survival. Damaged epithelial cells within the alveoli trigger a local immune response. A key immune cell is the macrophage, which can differentiate into a spectrum of phenotypes ranging from pro- to anti-inflammatory. To gain a greater understanding of the mechanisms of the immune response to VILI and post-ventilation outcomes in the absence of evolving comorbidities, we mathematically modeled interactions between the immune system and site of damage while accounting for macrophage phenotype. We generated a collection of parameter sets with biologically feasible dynamics and used statistical methods and sensitivity analysis to hypothesize predictors of outcome and interventions for poor response to ventilation. Additionally, we analyzed macrophage phenotype using a system of ordinary differential equations and an agent-based model, both of which focused on the spectrum of macrophage activation on an individual cell level. Using both platforms, we tested different scenarios to examine macrophage response to damage.
• Julia Arciero (Indiana University-Purdue University Indianapolis, USA)
  "Modeling novel immunoregulatory treatments for transplant patients"
Solid organ transplantation is a life-saving procedure that requires lifelong immunosuppression to prevent transplant rejection. Developing immunoregulatory treatments that minimize the need for chronic immunosuppression would be life-changing for transplant patients. Adoptive cell therapy with regulatory T cells (Treg) has emerged as a very promising approach, but there is limited understanding of the conditions that maximize Treg therapeutic effect. Mathematical modeling offers a unique and useful method for identifying cell therapy manipulations that would be most significant. This study introduces a mathematical model of transplant rejection that has been adapted to include adoptive transfer of Tregs with varied immunosuppression regimens. The model exhibits expected transplant behavior in the presence of immunosuppression, including graft acceptance with therapeutic levels of immunosuppression and graft rejection with subtherapeutic levels of immunosuppression. Preliminary results also indicate that combinatorial treatment strategies that incorporate adoptive transfer with subtherapeutic immunosuppression prolongs graft lifetime longer than either treatment in isolation. Ultimately, the model will be used to investigate optimal combinatorial dosing strategies that prevent graft rejection while minimizing immunosuppression. Modeling novel immunoregulatory treatments for transplant patients
• Josua Aponte-Serrano (Indiana University, USA)
  "Integrating Validated Models of Viral Replication and Interferon Signaling into a Multi-Scale Spatial Framework to Identify Key Factors of Viral Infection Dynamics"
Multi-scale models are commonly used tools to address complex problems that span over multiple biological scales: from intracellular signaling and regulatory pathways to host-level systemic responses. We present a multi-scale spatial model of RNA viral replication and type-I interferon response in epithelial cells. The parameters of the models were identified using using both in vivo and in vitro data from Influenza A Virus (IAV). We show that, by following our cellularization workflow, we can integrate independently validated models into a multi-scale framework that reproduces the dynamics of each model subcomponent. By exploring the parameter space of this integrated model we identified factors that lead to viral plaque growth arrest such as modulation of the JAK-STAT pathway and differential propagation of the interferon signal and viral particles in the extracellular environment. Sensitivity analysis of the integrated model suggest that parameters associated with the interferon signaling pathways are identifiable under experimental conditions that inhibit virus growth. Finally, we should how this multi-scale model can be extended to incorporate additional aspects of the host-immune response to viral infection.

Organized by: Jacques Bélair (Université de Montréal, Canada) & Elena Aruffo (York University, Canada)

• Matthew Betti (Mount Allison University)
  "Combining data forecasting with scenario-based modeling for insights into a rapidly changing outbreak situation"
We present a simple, modified SIR model with the intended use of bridging the gap betweeen data-fitted forecasts and modeled scenario-based forecasting. Using a combination of data-driven forecasting, simple model structures, and ensemble fitting we are able to determine mid-range predictions for rapidly changing situations. Using results over the past year on COVID-19 we will highlight the strengths of such an approach when it comes to forecasting trajectories and how this can be used to help policy and decision making.
• Marina Mancuso (Arizona State University)
  "Will Cross-Immunity Protect the Community Against COVID-19 Variants ?"
Several effective vaccines are currently being deployed to combat the COVID-19 pandemic (caused by SARS-CoV-2) around the world, resulting in significant reduction in the burden of the pandemic in places with high enough coverage. The effectiveness of COVID-19 vaccination programs is, however, significantly threatened by the emergence of new SARS-COV-2 variants that, in addition to being more transmissible and potentially more virulent than the resident strains, may at least partially evade existing vaccines. This talk is based on the use of a new multigroup and multi-strain mechanistic mathematical model for assessing the impact of the vaccine-induced cross-protective efficacy on the spread of the COVID-19 pandemic in the United States. In addition to estimating the vaccine-derived herd immunity threshold for the US, I will discuss conditions for which a new SARS-CoV-2 variant can fail to, or have the potential to, cause a significant surge in the US.
• Elena Aruffo (York University)
  "Vaccination rollout and relaxation of non-pharmaceutical interventions: a combined approach"
After months of implementation of non-pharmaceutical interventions to control the spread of SARS-CoV-2 infection, in December 2020 many countries began COVID-19 vaccination campaigns. Over the past few months, the vaccination coverage in Toronto, Canada increased visibly, leading to high immunization levels among certain age groups. In collaboration with Toronto Public Health, the Canadian Centre for Disease Modeling modeling group employed a deterministic structured compartmental model to investigate the current immunization status in Toronto and explore potential strategies for safe reopening, given various degrees of vaccine coverage by age group, in order to maximize reductions in cases, hospitalizations and deaths. We further examined the impact of different time intervals between the first and second vaccine dose on the aforementioned outcomes.
• Nicola Perra (University of Greenwich)
  "Modelling the COVID-19 pandemic at different spatio-temporal scales"
In the talk, I will provide an overview of different approaches I have applied to model the unfolding of the COVID-19 pandemic and its effects. In doing so, I will discuss the insights obtained by studying the initial phases of the pandemic, the first wave, and the vaccine rollout in the USA, Europe as well as Latin America. I will also discuss the key role of non-pharmaceutical interventions.

Organized by: Barun Majumder (University of Tennessee, USA), Soumen Bera (University of Tennessee, USA)
Note: this minisymposia has multiple sessions. The second session is MS18-IMMU.

• Joost Beltman (Division of Drug Discovery and Safety, Leiden Academic Centre for Drug Research, Leiden University, The Netherlands, The Netherlands)
  "Quantifying the role of T cells in tumor control through computational modeling"
Immunotherapies are an emerging strategy for treatment of solid tumors, for example by means of adoptive T cell therapies and stimulation of T cell functionality by specific antibodies. Improved understanding of the mechanisms employed by cytotoxic T lymphocytes (CTL) to control tumors will aid in the development of immunotherapies. CTLs can directly kill tumor cells in a contact- dependent manner or may exert indirect effects on tumor cells via secretion of cytokines. Here, we aim to quantify the importance of these mechanisms in various settings by application of computational models to experimental data acquired in mice. We developed ordinary differential equation models and agent- based models (ABMs) of tumor regression following adoptive transfer of a population of CTLs. Models were parameterized based on in vivo measurements of CTL infiltration over time, tumor volume measurements, and on image-based quantification of rates of tumor cell proliferation and apoptosis. We find that in two different settings direct, contact-dependent killing was insufficient to cause tumor regression and that antiproliferative effects by T-cell-produced cytokines have a large role in tumor control. Thus, our work highlights the potential importance of cytokine-induced antiproliferative effects in T-cell–mediated tumor control.
• Sachie Kanatani (Johns Hopkins Bloomberg School of Public Health, USA)
  "Comparative intravital imaging of human and rodent malaria sporozoites"
Malaria infection starts with the injection of Plasmodium sporozoites into the host's skin. Sporozoites are motile and move in the skin to find and enter blood vessels to be carried to the liver. We present the first characterization of P. falciparum sporozoites in vivo, analyzing their motility in mouse skin and human skin xenografts and comparing their motility to two rodent malaria species. These data suggest that in contrast to the liver and blood stages, the skin is not a species-specific barrier for Plasmodium. Indeed, P. falciparum sporozoites enter blood vessels in mouse skin at similar rates to the rodent malaria parasites. Furthermore, we demonstrate that antibodies targeting sporozoites significantly impact the motility of P. falciparum sporozoites in mouse skin. Though the sporozoite stage is a validated vaccine target, vaccine trials have been hampered by the lack of good animal models for human malaria parasites. Pre-clinical screening of next-generation vaccines would be significantly aided by the in vivo platform we describe here, expediting down-selection of candidates prior to human vaccine trials.
• Irina Grigorova (University of Michigan Medical School, USA)
  "Studying the role of CCL3 in the interactions between Germinal Center B cells and follicular regulatory T cells"
Follicular regulatory T cells (Tfrs) play multiple roles in the control of B cells response. From one side, they repress autoreactive and foreign antigen-specific germinal center (GC) B cells at the peak of GC response. From the other side, they promote GC B cell cycling in IL-10 dependent fashion and ensure optimal affinity maturation. However, which factors direct GC B cell interactions with Tfr cells has been unclear. Based on the single cell and bulk qPCR analysis we found that CCL3 is upregulated in about 10% of CCs that express Myc and are undergoing positive selection. Both ex vivo chemotaxis analysis and multiphoton intravital imaging suggests that CCL3 produced by GC centrocytes (CCs) promotes their direct contacts with Tfr cells. qPCR and transwell analysis revealed expression and synergistic involvement of CCR5 and CCR1 chemokine receptors on Tfr cells in their chemotaxis to CCL3. Both an adoptive transfer and mixed bone marrow chimeras models suggest that at the peak of GC response CCL3 promotes moderate repression of GC response. However, after the peak of GC response B cell-intrinsic production of CCL3 promotes prolonged participation of B cells in GCs, affinity maturation, as well as better memory and plasmablast response. To summarize, our studies suggest the existence of the local chemotactic cues between B cells and Tfr cells within GCs that direct interactions between the cells and are important for optimal regulation of GC response.
• Barun Majumder (Department of Microbiology, University of Tennessee Knoxville, USA)
  "Correlation between speed and turning naturally arises for sparsely sampled cell movements"
Mechanisms regulating cell movement are not fully understood. One feature of cell movement that determines how far cells displace from an initial position is persistence, the ability to perform movements in a direction similar to the previous movement direction. Persistence is thus determined by turning angles between two sequential displacements. Recent studies found that a cell's average speed and turning are negatively correlated, suggesting a fundamental cell- intrinsic program whereby cells with a lower turning ability (i.e., larger persistence) are intrinsically faster. Using simulations, we show that a negative correlation between the measured average cell speed and turning angle naturally arises for cells undergoing a correlated random walk due to sub- sampling, i.e., when the frequency of sampling is lower than frequency at which cells make movements. Assuming heterogeneity in persistence and intrinsic speeds of individual cells results in a negative correlation between average speed and turning angle that resembles experimentally observed correlations. Changing the frequency of imaging or calculating displacement of cohorts of cells with different speeds resulted in similar results whether or not there is a cell- intrinsic correlation between cell speed and persistence, and we could find many different parameter sets that allow to approximately match experimental data binned into cell cohorts. Interestingly, re-analysis of data of T cells in zebrafish showed that the observed correlation between persistence and speed is highly sensitive to sampling frequency, disappearing for coarsely sampled data. Our results thus challenge an established paradigm that persistent cells have intrinsically faster speeds and emphasize the role of sampling frequency may have on inference of critical cellular mechanisms of cell motility.

Organized by: Barun Majumder (University of Tennessee, USA), Soumen Bera (University of Tennessee, USA)
Note: this minisymposia has multiple sessions. The second session is MS17-IMMU.

• Paulus Mrass (Department of Molecular Genetics and Microbiology, University of New Mexico, USA)
  "Quantitative imaging identifies CXCR4 as a molecular switch that balances confinement and ballisitic migration of cytotoxic T cells within flu- infected lungs"
Cytotoxic T cells play an important role in protective immune responses against the flu, but the molecular mechanisms that regulate this function remain incompletely understood. In the present study we established a live imaging model that enables quantification of T cell motility within intact flu-infected lung tissue. This setup revealed that cytotoxic T cells show heterogenous migration patterns, characterized by intermittent periods of confinement and ballistic relocation. A special feature of our imaging model was the capacity to separately measure T cells that are in close proximity to flu-infected regions and those that are distant. Comparison of these two groups revealed that T cells that reside in flu-positive regions are signficantly more confined than T cells in flu-negative regions. This finding indicated that exposure to cognate peptides is one mechanism that contributes to the heterogeneous migration patterns of cytotoxic T cells within flu-infected lungs. To dissect the molecular mechanisms that regulate interstitial migration of T cells further, we analyzed T cell motility after treatment of lungs with pharmacological inhibitors. This approach revealed that AMD3100, a specific inhibitor of the chemokine recetpor CXCR4, caused a signficant suppression of interstitial migration within flu-negative regions. Unexpectely, we also found that inhibition of CXCR4 had an oppositive effect of T cells within flu-positive regions, i.e. the T cells became less confined. From these findings, we conclude that CXCR4 functions as a molecular switch that boosts interaction with target cells by two distinct mechanisms: (1) by enhancing motility towards flu-positive regions; and (2) by limiting motility within flu- positive regions, which likely facilitates the initiation of cognate interactions with target cells. Indeed, when we inhibited CXCR4 in flu-infected mice with AMD3100, this led to a reduction of degranulation of cytotoxic T cells infiltrating flu-infected lungs. Together, quantitative imaging has revealed that CXCR4 controls the functionality of lung-infiltrating cytotoxic T cells by regulation of intra-tissue motility.
• Arja Ray (Department of Pathology, University of California San Francisco, USA)
  "Visualizing T cell behavior in solid tumors to define barriers to immunotherapy"
Cancer immunotherapy relies on the effective function of cytotoxic CD8 T cells in the tumor microenvironment (TME). Other immune cells such as tumor- associated macrophages (TAMs) and the tumor stroma are critical components of the TME that inform CD8 T cell function. In tumors with abundant T cell infiltration, immunotherapy using bi-specific T cell engagers (BiTE) mediates physical interactions between T cells and tumor cells, thereby forcing tumor recognition and cytotoxic killing. However, this immunotherapy has had limited success in solid tumors, leading to questions regarding the barriers posed by the TME in this context. Using intravital imaging, we discovered vast heterogeneity in the movement of BiTEs out of perfused blood vessels in intact live tumors, from unhindered diffusion in some regions to being entirely contained within blood vessels in others. Indeed, the sufficiency of tumor-resident T cells to mediate tumor rejection was a function of dosage, thereby indicating that the bioavailability of such functional molecules in the TME is a key factor restricting their efficacy in solid tumors. Many solid tumors, on the other hand, are characterized by a lack of T cell (and other immune cell) infiltration, commonly referred to as an “immune desert” tumor. It has been postulated that TAMs play a key role in trapping T cells at the tumor margins, thereby leading to a T cell sparse tumor nest. Using a novel mouse model to specifically mark TAMs, we performed live imaging of TAM:T cell localization and interactions in the TME. Indeed, in an immune desert tumor model, T cells tend to be trapped near the tumor margins, co-localized with TAMs on a bed of robust deposition of fibrous collagen. Using spatial transcriptomics, we identify a unique TAM population at the tumor margin that are putatively involved in fibrosis in communication with CAFs. We hypothesize that this TAM subset is a key component of the immune- stromal cross-talk that leads to excessive fibrosis and exclusion of T cells from the TME in immune desert tumors. Overall, visualizing and defining the microenvironment around T cells in immune rich and immune desert tumors reveals distinct barriers to effective T cell function and points to the necessity of tailored approaches to improve cancer immunotherapy for different solid tumors.
• Judy Cannon (University of New Mexico School of Medicine, USA)
  "Effect of tissue environment on T cell movement"
T cells are a key effector cell type in the immune response, migrating through tissues in order to clear infection such as influenza infection in the lung. T cells must move through many different types of tissues to mount an effective response: naïve T cells migrate in and out of lymph nodes searching for antigen on dendritic cells, while activated T cells migrate to peripheral tissue such as lung to clear influenza infection. We investigate how different tissues such as lymph node and lung environments affect T cell motion using two photon microscopy to visualize effector T cells moving in different tissue settings. We perform quantitative analysis of in situ T cell movement and find that T cell speeds vary independent of the tissue environment or type of T cells. Naïve T cells in the lymph nodes move with similar average speed as effector T cells in the flu-infected lung, but effector T cells in an acute lung injury model move much more slowly. Interestingly, despite similar speeds, T cells in the lung do not show a coupling of speed and persistence that many other cell types have been seen to demonstrate, suggesting that the lung environment may exert effects on T cell movement to drive specific types of motion. T cells in the lung also show greater persistent motion than T cells in lymph nodes. The combination of in situ imaging and quantitative analysis of cell movement can uncover how specific tissue environments impact T cell movement and search for infection within different tissue contexts.
• Soumen Bera (Department of Microbiology, University of Tennessee Knoxville, USA)
  "Mathematical modeling of CD8 T cell-mediated elimination of malaria liver stages using intravital imaging experiments"
CD8+ T cells are one of the most critical immune defenses against intracellular pathogens capable of finding and eliminating the infected cells and preventing blood-stage diseases. Intravital imaging technic helps demonstrate the killing of liver stages Malaria parasites by memory induced or activated CD8+ T cells. Using these technics and mathematical modeling, we have recently shown the formation of large clusters consisting of variable number of effectors CD8+ T cells around the parasite-infected hepatocytes is rapid, indicating the high efficiency of CD8+ T cells for finding their target within complex organs like the liver. However, it has not been clear how many activated CD8+ T cells are required to eliminate the malaria parasites within a short period of time. Using a combination of intravital experimental data and mathematical modeling, we have provided detailed insights about the CD8+ T cells dynamic against the parasite phenotypes. The parasite's death corresponding to a high number of CD8+ T cells indicates a prolonged interaction between them; however, the death of parasites with a smaller number of T cells due to multiple factors. Using alternative mechanistic models, increasing the number of CD8+ T cells response better predict the parasite phenotypic dynamics compare to others, indicating increasing CD8+ T cells prompt the killing process. However, alternative mathematical models showed the fixed killing efficiency per T cell per parasite that means a higher number of T cells has higher killing efficiency. Finally, dose-response analysis indicates a smaller number of T cells is required to kill the parasites after a couple of hours of CD8+ T cells transfer, but with increasing time, a high number of T cells is required to eliminate the parasite. With different alternative methods, our analysis indicates novel insights about quantifying CD8+ T cells dynamic in the process of parasite elimination. 

Organized by: Luis Sordo Vieira (Division of Pulmonary, Critical Care, and Sleep Medicine, University of Florida, United States), Marissa Renardy (University of Michigan/Applied BioMath, United States), Tracy Stepien (Department of Mathematics, University of Florida, United States)
Note: this minisymposia has multiple sessions. The second session is MS20-IMMU.

• Borna Mehrad (Division of Pulmonary, Critical Care, and Sleep Medicine, University of Florida, United States)
  "Big Problems in Pulmonary Medicine: A Research Agenda"
According to the World Health Organization, 3 of the 10 leading causes of death worldwide are lung diseases. In order, these are pneumonia (in which category I include COVID-19 and tuberculosis), chronic obstructive pulmonary disease, and lung cancer — these illnesses are a good place to start a discussion about a research agenda about the big problems in pulmonary medicine. In this talk, I will give an overview of each illness from a clinical and biological perspective, discuss some recent discoveries in each field, and end with key unresolved questions for each category.
• Josh Mattila (University of Pittsburgh, United States)
  "Converting pathology into data points and back again: using systems immunology to investigate cause-effect relationships in tuberculosis"
Tuberculosis is caused by Mycobacterium tuberculosis (Mtb), a bacterium that infects nearly a third of the world’s population. The human immune system is very effective at combatting Mtb and most infected people never experience symptomatic TB but there are still more than 10 million new TB cases and almost 2 million deaths from TB per year. Granulomas are the hallmark of TB and these multicellular lesions form in Mtb-infected tissues. Under optimal conditions, granulomas prevent bacterial dissemination and can generate sterilizing immunity but under suboptimal conditions, granulomas are sites of bacterial persistence and replication. Unfortunately, it is difficult to identify correlates of immunity in TB because granulomas occur in tissues that cannot be sampled and most of our information on immunity in TB comes from peripheral blood or murine TB models, neither of which replicate fully immunity in granulomas. Granulomas from experimentally-infected nonhuman primates (NHP) offer a human-like alternative but inter-granuloma heterogeneity and difficulties assessing the temporal trajectory of granuloma maturation and function make it difficult to interpret data from NHP granulomas. Computational models of granulomas, powered by biologic data obtained from ex vivo wet-lab studies on NHP granulomas, can model aspects of granuloma biology that correlate with protective or detrimental immunity. Here, I review how we have used biologic data from NHP granulomas to calibrate and validate GranSim, a computational granuloma model developed by the Kirschner Lab at the University of Michigan.
• Maral Budak (University of Michigan Medical School, United States)
  "Optimization of multidrug therapies for tuberculosis using a multi-scale computational model"
Tuberculosis (TB) is caused by the inhalation of Mycobacterium tuberculosis (Mtb), leading to ~1.5 million deaths every year. Mtb mainly infects lungs and triggers the formation of dense cellular structures composed of immune cells, bacteria, and dead tissue, called granulomas. The complex structure of granulomas prevents the effective penetration of antibiotics used to treat TB. Moreover, the heterogeneity of granulomas gives rise to various microenvironments for Mtb, where bacteria acquire different metabolic states that determine the potency of antibiotics either singly or in combination. Due to these reasons, TB treatment requires treatment with multiple antibiotics over long periods (6-9 months), causing prolonged side effects and compliance issues. Optimizing multidrug therapies and regimens for TB is essential to treat TB more effectively. In this study, we combined in vitro drug interaction predictions within GranSim, our computational model of granuloma formation and drug activity that simulates spatio-temporal granuloma drug dynamics. By systematically testing drug candidate regimens and considering drug interactions, we predict optimal drug regimens to be tested in vivo. This study will potentially lead to the discovery of more effective drug regimens that shorten the treatment window and have fewer side effects.
• Henrique de Assis Lopes Ribeiro (Division of Pulmonary, Critical Care, and Sleep Medicine, University of Florida, United States)
  "Computational Modeling Reveals the Role of Macrophages in Respiratory A. fumigatus Infection in Immunocompromised Hosts"
Fungal infections of the respiratory system are a life-threatening complication for immunocompromised patients. Invasive pulmonary aspergillosis, caused by the airborne mold Aspergillus fumigatus, has a mortality rate of up to 50% in this patient population. The lack of neutrophils, a common immunodeficiency caused by, e.g.,chemotherapy, disables a mechanism of sequestering iron from the pathogen, an important virulence factor. This paper shows that a key reason why macrophages are unable to control the infection in the absence of neutrophils is the onset of hemorrhaging, as the fungus punctures the alveolar wall. The result is that the fungus gains access to heme-bound iron. At the same time, the macrophage response to the fungus is impaired. We show that these two phenomena together enable the infection to be successful. A key technology used in this work is a novel dynamic computational model used as a virtual laboratory to guide the discovery process. The paper shows how it can be used further to explore potential therapeutics to strengthen the macrophage response.

Organized by: Luis Sordo Vieira (Division of Pulmonary, Critical Care, and Sleep Medicine, University of Florida, United States), Marissa Renardy (University of Michigan/Applied BioMath, United States), Tracy Stepien (Department of Mathematics, University of Florida, United States)
Note: this minisymposia has multiple sessions. The second session is MS19-IMMU.

• Julie Leonard-Duke (University of Virginia/Robert M. Berne Cardiovascular Research Center, United States)
  "Computational Modeling of Fibroblast Subpopulations in Idiopathic Pulmonary Fibrosis"
Each year in this country, 40,000 patients are diagnosed with idiopathic pulmonary fibrosis (IPF), a progressive and terminal disease caused by excessive extracellular matrix production by fibroblasts in distributed lesions, or “fibrotic foci”, throughout the lung. Fibroblasts are the primary pathologic cell population in fibrosis and their presence has been shown to be essential for fibrotic foci formation. Their actions, such as proliferating, secreting collagen, or differentiation into myofibroblasts, is driven by a combination of mechanical and chemical cues that eventually lead to a pathologic phenotype in IPF. Recent literature suggests that there are sub-populations of fibroblasts in the lung that exhibit different phenotypes depending on chemical and mechanical signals present in their local environment. Understanding how fibroblast phenotypic heterogeneity contributes to fibrotic foci formation in the dynamic lung environment of progressive IPF is an overarching goal of our research team and has important implications in the design of new therapies for IPF. Our group has recently performed single-cell RNAseq analysis on human lung fibroblasts exposed to a combination of pro-inflammatory cytokines to recapitulate the IPF lung environment. This analysis has led to the identification of fibroblast sub-populations that may behave differently from one another in response to their local and changing environment. To better understand the consequences of these phenotypic differences on lung tissue remodeling, our team is combining data-driven analyses with multi-scale agent-based modeling that simulates intracellular signaling and multi-cell interactions to predict cell-specific behaviors that are crucial to the formation of fibrotic foci in IPF.
• Amber M. Smith (University of Tennessee Health Science Center, USA)
  "Bacterial coinfections: from influenza to SARS-CoV-2"
Influenza virus infected individuals often become coinfected with a bacterial pathogen, which significantly enhances morbidity and mortality. These bacterial coinfections have contributed to 45-95% of mortality during influenza pandemics, and numerous host and pathogen mechanisms have been identified through various experimental and mathematical modeling approaches. Given the history of influenza-bacterial coinfections, this was an obvious fear for the ongoing SARS-CoV-2 pandemic. Thus far, there is some evidence that SARS-CoV-2 also increases susceptibility to bacterial infections but does so to a lesser extent compared to influenza. To better understand the potential for SARS-bacteria coinfection, we infected mice with SARS-CoV-2 followed by pneumococcus. Our data support clinical observations and highlight specific host responses that may play a role in the increased pathogenicity.
• Elsje Pienaar (Purdue University, United States)
  "Mycobacterium avium infection in the lungs: and agent-based model exploring early infection events"
INTRODUCTION: Mycobacterium avium complex (MAC), members of the nontuberculous mycobacteria family, are environmental microbes, capable of colonizing and infecting humans following inhalation of the bacteria. MAC-pulmonary disease is notoriously difficult to treat and prone to recurrence, and both incidence and prevalence have been increasing [1]. There are two types of MAC lung infection – fibrocavitary and nodular, with fibrocavitary much harder to treat, and with much lower cure rates, as low as 76% even with optimal treatment [2]. MAC are well known to form biofilms and diverse colonies in the environment. These biofilms have been shown to aid in epithelial cell invasion [3], cause premature apoptosis in macrophages [4], and inhibit antibiotic efficacy [5]. We hypothesize that both phenotypic diversity and biofilm formation are key to establishing and prolonging infections in the lung. To address these hypotheses, we developed a model that shows the interactions between bacteria, biofilm and immune cells as an agent-based model (ABM). This model allows us to explore both the intracellular scale (bacterial phenotypes and macrophage killing), and tissue scale (biofilm formation and epithelial invasion). METHODS: We used Repast Simphony to develop a three-dimensional ABM of in vivo MAC colonization to infection within the first 14 days post-deposition. The grid represents a length of lung airway with a layer of mucus/epithelial lining fluid (ELF). Bacteria agents are divided into either sessile (slow-growing, within biofilm and less susceptible to antibiotics), or planktonic (more quickly growing but not protected by biofilm) phenotypes. Biofilm is represented by continuous variables in each grid compartment, with values corresponding to the amount of extracellular matrix produced by bacteria in that grid compartment. To represent the protective properties of biofilm, the amount of biofilm is inversely related to the likelihood of a macrophage phagocytosing bacteria from that biofilm. All bacterial agents also release a chemoattractant that is represented by continuous variables in each grid compartment, and that diffuses throughout the grid. Macrophages probabilistically follow this chemoattractant gradient. Macrophages can phagocytose bacteria, prioritizing planktonic bacteria (not within biofilms), which infect the macrophage. Infected macrophages then have a probabilistic chance of killing internal bacteria. Macrophages also accumulate “apoptotic signal” through exposure to biofilm and internal bacteria. RESULTS: The model was parameterized through a literature search, test cases based on in vitro experiments and Latin Hypercube Sampling for unknown parameter values. We found that parameters affecting macrophage chemotaxis and recruitment have significant impact on the number of macrophages, but not on the number or distribution of bacteria. Initial parameters – the initial bacteria count, initial macrophage count, and ratio of planktonic to sessile bacteria - have lasting impacts throughout the simulation. Parameters that pertain to only one bacterial subpopulation (e.g. extracellular growth rates) are not significantly correlated with outcomes overall, because the composition of the bacterial populations varies so much between simulations. Finally, we have found that biofilm increases the number of bacterial cells that invade the epithelium, but in the absence of biofilm bacteria are able to persist in the airways. Higher biofilm levels also increase macrophage chemo-attractant production, death and recruitment. The most significant biofilm parameter is the amount that is deposited with bacteria in the lung upon initial exposure. Our simulations indicate that, based on in vitro data, once bacteria are deposited in the lung they cannot generate biofilm quickly enough to have a significant an impact. CONCLUSIONS: We have developed a multiscale agent-based model that allows us to study the initial colonization and infection in MAC-pulmonary disease on both the cellular- and tissue level. Early results show that initial parameters have lasting effects on the outcome of the deposition. Further, we have found that biofilms are not necessary to establish fibrocavitary type of MAC infection. Future directions of this work include organization of the infection into nodules, adding drug pharmacokinetics and pharmacodynamics to better understand the role of bi¬¬ofilm in treatment efficacy. REFERENCES: 1. Lee, et al. Antimicrob Agents Chemother, 59(6): 2972-2977, 2015. 2. Hwang, et al. Eur Respir J, 49(3): 2017. 3. Yamazaki, et al. Cell Microbiol, 8(5): 806-814, 2006. 4. Rose and Bermudez. Infect Immun, 82(1): 405-412, 2014. 5. Falkinham. J Med Microbiol, 56(Pt 2): 250-254, 2007.
• Angela Reynolds (Virginia Commonwealth University, United States)
  " Mathematical modeling of lung inflammation from insult to recovery"
Lung inflammation can be triggered by many insults including viral and bacterial infections, structural damage, or inhalation of dangerous particles. The associated lung injury can resolve quickly, be treated effectively through various interventions, become a chronic problem, or lead to death. Given the variety of possible responses, often seen from the same insult, and the necessity for the lungs to function effectively mathematical modeling has become a necessary tool for improving lung health. Researchers have used mathematical modeling to understand immune system dynamics during a number of pulmonary infections and injuries, identify key mechanisms, and provide important insights into new treatments and to help identify who needs an intervention. In this talk we will review and explore recent research in mathematical modeling of inflammation in the lung and look into how mathematical modeling and computational methods can be used to guide interventions.

• Aaron Meyer University of California, Los Angeles
  "Developing a mechanistic view of mixed IgG antibody immune effector responses"
IgG antibodies bind antigen targets and then interact with Fcγ receptors (FcγR) on effector cells to direct cellular responses. Effector responses involve multiple cell types and processes (e.g., cytokines, phagocytosis) making it a systems-level challenge to precisely engineer these responses. We previously showed a multivalent binding model could accurately predict in vitro binding to synthetic complexes and in vivo anti-tumor antibody response.Here, we extend this work to predict the binding and immune response to complexes with combinations of Fc domains using a multivalent, multi-receptor, and multi-ligand model. We first validated the accuracy of this model through binding experiments using synthetic IgG mixtures. Applying this model, we could predict the effector-elicited target depletion of individual IgG and their combinations in mouse models of both ITP and melanoma. Our model correctly inferred that Kupffer cells are essential for platelet depletion in ITP, and specifically identified a FcγRIIBhigh subpopulation with outsized importance. Exploring the predicted effects of IgG combinations, we develop a framework for drug additivity. IgG synergy cannot occur with antibodies of identical antigen binding but antagonism is widespread through Fc receptor competition. These results demonstrate a suite of capabilities to more precisely engineer antibodies.
• Yafei Wang Indiana University Bloomington
  "Multiscale modeling of SARS-CoV-2 infectious dynamics and antiviral drugs intervention"
The 2019 novel coronavirus, SARS-CoV-2, is a pathogen of critical significance to international public health. Antiviral drugs and vaccines are currently under development and test to address this pandemic. However, the knowledge of dynamics of viral endocytosis, replication, single-cell response and pharmacodynamics is still limited. Therefore, the establishment of such dynamics is very urgent for understanding how SARS-CoV-2 infectious spread and finding an optimal therapeutic strategy. More importantly, this pandemic will likely not be the last one as new pathogens emergency, so we may be able to reuse the dynamics for faster response to another pandemic in the future. In this talk, we will present our work on multiscale modeling of SARS-CoV-2 infectious spread and antiviral drugs intervention (through an agent-based modeling approach-PhysiCell). This work is an extension of a multi-institution, multi-disciplinary coalition of over 40 mathematical biologists, immunologists, virologists, pharmacologists, and others to build a comprehensive multiscale model of SARS-CoV-2 infection dynamics and immune response. The model of this work can be run on a cloud-hosted platform at: https://nanohub.org/resources/virion2pbpd.
• Wenjing Zhang Texas Tech University
  "Deterministic and Stochastic in-host Tuberculosis Models for Bacterium-directed and Host-directed Therapy Combination"
The goal of this paper is to investigate in-host tuberculosis models to provide insights into therapy development. Focusing on therapy-targeting parameters, the parameter regions for different disease outcomes are identified from an established ODE model. Interestingly, the ODE model also demonstrates that the immune responses can both benefit and impede disease progression, which depend on the number of bacteria engulfed and released by macrophages. We then develop two Ito SDE models, which consider demographic variations at the cellular level only and environmental variations during therapies along with demographic variations. The SDE model with demographic variation suggests that stochastic fluctuations at the cellular level have significant influences on (1) the T-cell population in all parameter regions, (2) the bacterial population when parameters locate in the region with multiple disease outcomes, and (3) the uninfected macrophage population in parameter region representing active disease. Further, considering environmental variations from therapies, the second SDE model suggests that disease progression is more likely to be inhibited if therapies (1) can have fast return rates and (2) are able to bring parameter values into the disease clearance regions.
• Aniruddha Deka Pennsylvania State University
  "Pathogen competition and mutant invasion in face of human choice in vaccination:"
Competition between multiple strain for vaccine preventable diseases often leads to exclusion of some pathogens, while it may influence the invasion of an emerging mutant in the population. Previous studies have shown that basic reproductive numbers among multiple strains are sufficient to predict which strains will invade a population. But human vaccination decision plays crucial role in shaping the type of strain that will invade or persist or get eliminated. Humans adapt to changing behavior or virulence of strains and for highly transmissible strains, they vaccinate at a faster rate due to higher perceived severity from the diseases. This on the other-hand gives scope for mutant strains to invade new number of susceptible in the population. In our study, we have coupled game dynamic model of vaccination choice and compartmental disease transmission model of two-strains to explore invasion, extinction and persistence of a mutant in the population which have a lower reproduction rate than the resident one. We illustrate that higher perceived strain severity and lower perceived vaccine efficacy are necessary conditions for persistence of a mutant strain. Numerically we explore these invasion and persistence analyses under asymmetric cross-protective immunity of these strains.

• Girma Mesfin Zelleke AIMS-Cameroon
  "A Mathematical Understanding of the Dynamic Regulation of the Complement System on bacterial infection."
The innate immune system responds to bacterial infections first by activating the fastest defense mechanism called the complement system. This system is controlled by more than 30 different proteins that work as a team to damage the invader and to alert other immune systems. However, this defense mechanism is perilous if the activation is abnormal, inefficient, and overstimulated, or if there are deficiencies in a surface-bound with receptors of the invaders. It is therefore vital that the complement system binds with an invading bacterium successfully allowing the cascade of events that will enable a proper stimulation of the defensive mechanism by the complement system. Here, we propose a mathematical model which describes the dynamics of the complement system against bacterial infection. We further investigate the mathematical and numerical analysis of the model which generates and explains conditions for normal, efficient, and properly stimulated concentration of the complement system proteins. We also perform sensitivity analysis to identify the critical parameters that affect the direct action of the complement system on the bacterial infection.
• Solveig A. van der Vegt Wolfson Centre for Mathematical Biology, Mathematical Institute, University of Oxford, UK
  "Mathematical modelling of autoimmune myocarditis and the effects of immune checkpoint inhibitors"
Autoimmune myocarditis, or inflammation of cardiac muscle tissue, is a rare but potentially fatal side effect of cancer treatment with immune checkpoint inhibitors. Of patients receiving this type of treatment, approximately 1% develop myocarditis, and the disease proves to be fatal in about 25-50% of these cases. Despite the severity of this side effect and the large volume of cancer patients eligible for treatment with immune checkpoint inhibitors, no preclinical assay currently exists that tests new compounds for myocarditis-related cardiotoxicity. Our aim is to use mathematical modelling to develop a better understanding of the immune cell types and mechanisms involved in the development and progression of autoimmune myocarditis and the effects that immune checkpoint inhibitors have on this. To this end, we have developed the first mathematical model of this disease. By employing parameter sensitivity methods and examining the bifurcation structures in the model, we aim to pinpoint the critical cell types that have to be included in the preclinical test for it to reflect well the mechanisms involved in the development of drug-induced autoimmune myocarditis in vivo.
• Martin Lopez-Garcia Department of Applied Mathematics, University of Leeds
  "A stochastic multi-scale model of Francisella tularensis infection"
We present a multi-scale model of the within-phagocyte, within-host and population-level infection dynamics of Francisella tularensis, which extends the mechanistic one proposed by Wood et al. (2014). Our multi-scale model incorporates key aspects of the interaction between host phagocytes and extracellular bacteria, accounts for inter-phagocyte variability in the number of bacteria released upon phagocyte rupture, and allows one to compute the probability of response, and mean time until response, of an infected individual as a function of the initial infection dose. A Bayesian approach is applied to parameterize both the within-phagocyte and within-host models using infection data. Finally, we show how dose response probabilities at the individual level can be used to estimate the airborne exposure to Francisella tularensis in indoor settings (such as a microbiology laboratory) at the population level, by means of a deterministic zonal ventilation model.
• Mohammad Aminul Islam University at Buffalo, The State University of New York
  "Modeling the Progression of Fibrosis with Dysregulation of ACE2 in COVID19 Patients"
The severity of the COVID19 pandemic creates an emerging need to investigate the long-term effect of infection on healing patients. Many individuals are at the risk of suffering pulmonary fibrosis due to pathogenesis of lung injury and impairment in the healing mechanism. SARS-CoV-2 enters the host cells via binding its spike protein with the ACE2 receptor which is a key component in modulating the balance of the renin-angiotensin system (RAS). The dysregulation of ACE2 by the viral infection can shift the balance of RAS towards pro-inflammation and pro-fibrosis. We developed a multiscale agent-based model to investigate the dynamics of viral infection, immune cell response, and fibrosis in lung tissue. The model can simulate the dynamics of ACE2 and collagen deposition in the 2D lung tissue at different severity of infections. We use the ACE2 dynamics as input in a separate model of RAS to predict the change in ANGII, which is a mediator for pro-inflammation and collagen deposition which is a mediator for pro-fibrosis from homeostasis for normotensive and hypertensive patients. Our model also reveals that the variation in available ACE2 due to age and gender can lead to significant change in inflammation, tissue damage, and fibrosis.

• Gulsah Yeni Pennsylvania State University
  "Modeling PrEP on Demand for Prevention of HIV"
In order to prevent the spread of HIV, antiretroviral therapy (ART) for HIV drugs can be administered to high-risk individuals in advance of exposure, as pre-exposure prophylaxis (PrEP). PrEP with the ART combination drug Truvada taken daily has been demonstrated to effectively reduce the risk of HIV infection. However daily dosing can be onerous, and studies suggest that short-term use of ARTs around the time of exposure may be just as effective at reducing HIV risk. Here we investigate such “on-demand” PrEP. We build a mathematical framework in which we integrate a pharmacokinetic/pharmacodynamic (PK/PD) model developed by measuring mucosal tissue concentrations of tenofovir and emtricitabine (Truvada) (Cottrell et al. 2016) into an in-host stochastic model of early HIV infection with PrEP treatment based on virus dynamics. Armed with this model, we predict risk of infection under different on-demand PrEP regimens with regards to time of dosing and dosage relative to time of exposure. Thus we predict practical on-demand PrEP regimens in terms of dosage and timing required to obtain most effective protection for lower female genital tract (FGT).
• Esteban Abelardo Hernandez Vargas UNAM
  "Topological Data Analysis in Infectious Diseases"
Pathogens have important implications in many aspects of health, epidemiology, and evolution. Topological Data Analysis (TDA) is used here to help in identifying the behaviour of a biological system from a global perspective. Using data sets of the immune response during influenza-pneumococcal co-infection in mice, we employ here topological data analysis to simplify and visualise high dimensional data sets. Persistent shapes of the simplicial complexes of the data in the three infection scenarios were found: single viral infection, single bacterial infection, and co-infection. The immune response was found to be distinct for each of the infection scenarios and it was uncovered that the immune response during the co-infection has three phases and two transition points.
• Laura Liao Merck & Co., Inc
  "BUILDING A MECHANISTIC MODELING PLATFORM FOR HIV CURE DRUG DEVELOPMENT"
Current antiretroviral therapy (ART) effectively controls HIV in most patients but does not cure it. To develop drugs towards a HIV cure, novel approaches – such as reactivation of latent provirus (“shock”) and immunotherapies – are being explored. Mechanistic mathematical models that describe both within-host viral load dynamics and immunologic control of HIV infection are essential to integrate clinical data, assess therapeutic response, and generate hypotheses in support of HIV cure drug development. We built the Immune Viral Dynamics Modeling (IVDM) platform, based on recently developed mathematical models which integrate potential mechanisms that may lead to a cure. To inform the IVDM parameters, we created a dataset of “artificial” subjects by concatenating post-ATI (analytical treatment interruption) viral load profiles from eight ACTG clinical studies with on-ART viral load from a clinical study of raltegravir. This way, we created a dataset that describes the course of infection from initiation of treatment to ATI. Key parameters that govern latent reservoir seeding and immunological control were estimated using a nonlinear mixed effects approach (Monolix). From the estimated parameter distributions, we sampled a virtual population and ran clinical trial simulations (CTS) to assess potential curative interventions.

• Ke Li The University of Melbourne
  "Modelling the effect of MUC1 on influenza virus infection kinetics and macrophage dynamics"
The host immune response is important to defend against influenza viral infection. However, overstimulation of the host immune response can lead to pathology, indicating a subtle balance between a protective and a destruct response. Dysregulated immune responses are often associated with an excessive recruitment of macrophages. MUC1 belongs to the family of cell surface (cs-) mucins and has been shown to be an important and dynamic component of the host innate immune response, associated with recruitment of macrophages. Experimental evidence indicates that its presence reduces influenza infection severity. However, the detailed effects of MUC1 in vivo remain elusive, limiting our ability to predict the efficacy of potential treatments that target MUC1. To address this limitation, we fit two mathematical models to available in vivo kinetic data for both virus and macrophage populations in wild-type and MUC1 knockout mice. Both models provide evidence that MUC1 reduces the susceptibility of epithelial cells and show that the MUC1 regulates the recruitment of macrophages and thus the host immune response. This study improves our understanding of the dynamic role of MUC1 against influenza infection and may support the development of novel antiviral treatments.
• Juan Antonio Magalang Theoretical Physics Group, National Institute of Physics, University of the Philippines
  "Stochastic resetting antiviral therapies prevent drug resistance development"
We study minimal mean-field models of viral drug resistance development in which the efficacy of a therapy is described by a one-dimensional stochastic resetting process with mixed reflecting-absorbing boundary conditions. We derive analytical expressions for the mean survival time for the virus to develop complete resistance to the drug. We show that the optimal therapy resetting rates that achieve a minimum and maximum mean survival times undergo a second- and first-order phase transition-like behaviour as a function of the therapy efficacy drift. We illustrate our results with simulations of a population dynamics model of HIV-1 infection.

• Ellie Mainou Pennsylvania State University
  "Investigating model alternatives for acute HIV infection"
The standard viral dynamics model explains HIV viral dynamics during acute infection reasonably well. However, the model makes simplifying assumptions, neglecting some aspects of HIV pathogenesis. For example, in the standard model, target cells are infected by a single HIV virion. Yet, cellular multiplicity of infection (MOI) may have considerable effects in pathogenesis and viral evolution. Further when using the standard model, we take constant infected cell death rates, simplifying the dynamic immune responses. Here, we use four models—1) the standard viral dynamics model, 2) an alternate model incorporating cellular MOI, 3) a model assuming density-dependent death rate of infected cells and 4) a model combining (2) and (3)—to investigate acute infection dynamics among study participants in the RV217 dataset. We find that all models explain the data, but different models describe differing features of the dynamics more accurately. For example, while the standard viral dynamics model may be the most parsimonious model, viral peaks are better explained by a model allowing for cellular MOI. These results suggest that heterogeneity in within-host viral dynamics cannot be captured by a single model but depending on the aspect of interest, a corresponding model should be employed.
• Christian Quirouette Ryerson University
  "Time to revisit the endpoint dilution assay"
A virus sample's infectivity is measured by the number of the infections it causes per unit volume, via a plaque or focus forming assay (PFU or FFU) or an endpoint dilution (ED) assay (TCID50, EID50, etc.). The plaque and focus assays have several technical and experimental limitations we will outline in this presentation, but yield a simple measure: one plaque equals one infectious dose. The ED assay does not suffer from these limitations, but as we will show, the measure it yields, the TCID50, is calculated using biased and antiquated approximations that relate poorly to the number of infectious doses in the sample. We propose taking the best of both: (1) preferring the ED assay over the more subjective plaque or focus forming assay; and (2) replacing the TCID50 with an accurate, robust and meaningful measure we call Specific INfections or SIN, corresponding to the most likely number of infections a virus sample will cause. We will demonstrate how the measure of SIN compares to current measures (FFU, TCID50) under typical experimental conditions, and how experimental protocols can be altered to yield even more accurate measures.
• Bevelynn Williams University of Leeds
  "A stochastic intracellular model of anthrax infection with spore germination heterogeneity"
During inhalational anthrax infection, Bacillus anthracis spores are ingested by alveolar macrophages, and begin to germinate and then proliferate inside them, which may eventually lead to death of the host cell and the release of bacteria into the extracellular environment. Alternatively, some macrophages may be successful in eliminating the intracellular bacteria and will recover. In this talk, we consider a stochastic model of the intracellular infection dynamics of B. anthracis in macrophages. We explore the potential for heterogeneity in the spore germination rate, with the consideration of two extreme cases for the rate distribution: continuous Gaussian and discrete Bernoulli. This model has been calibrated by means of approximate Bayesian computation, using experimental measurements. We use the calibrated stochastic model to predict the probability of rupture, mean time until rupture, and rupture size distribution, of a macrophage that has been infected with one spore. We also obtain the mean spore and bacterial loads over time for a population of cells, each assumed to be initially infected with a single spore. Our results support the existence of significant heterogeneity in the germination rate across different spores, with a subset of spores expected to germinate much later than the majority.
• Barbara Szomolay Cardiff University
  "Computational Identification of Cancer Immunotherapy Targets using Combinatorial Peptide Libraries"
The interaction between T-cell receptors (TCRs) and peptides is highly degenerate: a single TCR may recognize about one million different peptides in the context of a single MHCI molecule. On the other hand, TCR recognition is fundamentally peptide- and/or MHC-specific: the functional sensitivity, which can be viewed as experimental realisation of the TCR triggering rate, is large enough only for minute fraction of all possible ligands. TCR triggering rate and degeneracy are mathematical concepts that are fundamental for an approach that uses length-matched combinatorial peptide library (CPL) scan data to search protein databases and to rank peptides in order of likelihood recognition. This CPL-based database screening can, to a large extent, accurately identify self-peptides that triggered the CD8 T-cell. The computational time required for peptide searching can be significantly reduced by using graphics processing units (GPUs). Adoption of GPU-accelerated prediction of T-cell agonists has the capacity to revolutionise our understanding of cancer immunity by identifying potential targets for tumor-specific T-cells.

• Macauley Locke University of Leeds
  "Novel Stochastic Models of type 1 interferon inhibition by Ebola Virus VP35"
The 2014 West Africa Ebola virus (EBOV) epidemic resulted in an increased desire and urgency to identify the mechanisms explored by EBOV to subvert immune responses; in particular that of type I interferon, which is a prototype innate immune response to a viral infection. This family of cytokines is important in the early stages of infection and key to inducing antiviral states within infected cells. There exists ample experimental evidence of the role that the EBOV 'multi-function'' protein VP35 has in promoting antagonism in a number of antiviral signalling pathways. We have developed novel stochastic models of VP35 antagonismin the type I interferon induction pathway based on current empirical evidence.Making use of approximate Bayesian computation, the mathematical modelsand experimental data sets, we have carried out model selection (to test differentmolecular hypotheses) and parameter calibration. Experimentaldata from the EBOV animal model of in vivo infection of rhesus monkeys.With a wish to gain further understanding into early time dynamics of type I interferon production during EBOV infections (or other viral infections, such as SARS-CoV-2), we hope that these models can be further extended to other viruses and their methods of innate immuneinhibition.
• Christopher Rowlatt University of St Andrews
  "Modelling the within-host spread of SARS-CoV-2 infection, and immune response, using a multi-scale individual-based model"
The COVID-19 pandemic, caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has affected millions of people worldwide. Although the majority of cases present asymptomatic or mild symptoms that do not require hospitalisation, many cases can develop into severe disease (such as acute respiratory disease syndrome (ARDS)) requiring hospitalisation, ventilation and may result in death. A hyper-active or dysfunctional immune response (such as increased monocyte/macrophage and neutrophil infiltration) coupled with an excessive pro-inflammatory cytokine response (such as high levels of interleukin-6) are believed to play a prominent role in the development of severe disease. However, the precise mechanisms that lead to severe disease remain unclear. In this talk, we employ a hybrid multi-scale individual-based model to study the spread of SARS-CoV-2 on an epithelial monolayer. We focus our attention on the early dynamics of the host innate immune response and the immune cell cross-talk, as well as the interaction with secreted cytokines.
• Giulia Belluccini University of Leeds
  "Multi-stage model of cell proliferation and death"
Many biological processes are modelled using Markov chains; thus, the inter-event times are assumed to be exponentially distributed. This hypothesis fails when cell proliferation plays a key role in the system of interest. Indeed, the history-dependence nature of the cell cycle breaks the Markov property. Our description of a population of stimulated cells preserves some of the convenient properties of a Markov process. A cell's time to division is a random variable with an Erlang distribution; its time to death has an exponential distribution. The underlying idea is to divide the cell cycle into a number of stages, each exponentially distributed and independent of the others. We can then consider cell generations in the model; that is all the cells that have divided the same number of times. Hence, the number of divisions that the cell has undergone is tracked, and this makes the model parameterisation feasible. The parameters are inferred using an Approximate Bayesian Computation approach based on sequential Monte Carlo methods with CFSE data of human and murine T lymphocytes. We calibrated the exponential model and the multi-stage one, and compared them using the corrected Akaike Information Criterion to prove statistically the better fit of the multi-stage model.
• Daniel Luque Duque University of Leeds
  "Multivariate competition for survival stimulus of T cell clonotypes in homeostasis"
A mechanism used to maintain the naive T cell repertoire is competition for homeostatic proliferation stimuli. We propose a multivariate competition model to study the dynamics of n clonotypes competing for proliferation stimulus provided by self peptides bound to Major Histocompatibility Complexes (self pMHC). We assume the population of self pMHC to be at steady state (thus providing constant stimulus) and study the long-term behaviour of the system by analysing: (i) dynamics in the long-term before extinction through the quasi-stationary distribution; (ii) time to extinction of the first clonotype, probabilities of extinction for each clonotype, and size of the surviving ones when extinction occurs. Additionally we analyse the distribution of the number of divisions of a given clonotype before its extinction.