Search

This paper describes and validates an algorithm to solve optimal control problems for agent-based models (ABMs). For a given ABM and a given optimal control problem, the algorithm derives a surrogate model, typically lower-dimensional, in the form of a system of ordinary differential equations (ODEs), solves the control problem for the surrogate model, and then transfers it back to the original ABM. It applies to quite general ABMs and offers several options for the ODE structure, depending on what information about the ABM is to be used. There is a broad range of applications for such an algorithm, since ABMs are used widely in the life sciences, such as ecology, epidemiology, and biomedicine and healthcare, areas where optimal control is an important purpose for modeling, such as for medical digital twin technology. Author summary: The motivation for the work reported in this paper is the development of mathematical tools for medical digital twin development. Based on a computational model of some aspects of human biology, there is a two-way interaction between the physical twin (the patient) and the digital twin (the model). In one direction, the model is periodically calibrated with patient-derived data to evolve together with the patient, making the model into a digital twin, and in the other direction, optimal interventions derived from the digital twin are administered to the patient. In many cases, the underlying computational model does not readily provide optimal control methods to identify interventions. Model types such as agent-based models (ABMs) are often more suitable than models consisting of ordinary differential equations (ODEs). In this paper, we present an algorithm that takes as input a general ABM, together with an optimal control problem and provides as output a solution. This is accomplished by first constructing a surrogate ODE model, solving the optimal control problem, and then lifting it to the ABM. The algorithm provides for several different types of surrogate models, ranging from those that implement mechanistic features of the ABM to purely phenomenological models. The algorithm is validated by applying it to a predator-prey ABM and a metabolic network represented as an ABM. [ABSTRACT FROM AUTHOR]

Trehalose uptake at 65[degrees]C in Rhodothermus marinus was characterized. The profile of trehalose uptake as a function of concentration showed two distinct types of saturation kinetics, and the analysis of the data was complicated by the activity of a periplasmic trehalase. The kinetic parameters of this enzyme determined in whole cells were as follows: [K.sub.m] = 156 [+ or -] 11 [micro]M and [V.sub.max] = 21.2 [+ or -] 0.4 nmol/min/mg of total protein. Therefore, trehalose could be acted upon by this periplasmic activity, yielding glucose that subsequently entered the cell via the glucose uptake system, which was also characterized. To distinguish the several contributions in this intricate system, a mathematical model was developed that took into account the experimental kinetic parameters for trehalase, trehalose transport, glucose transport, competition data with trehalose, glucose, and palatinose, and measurements of glucose diffusion out of the periplasm. It was concluded that R. marinus has distinct transport systems for trehalose and glucose; moreover, the experimental data fit perfectly with a model considering a high-affinity, low-capacity transport system for trehalose ([K.sub.m] = 0.11 [+ or -] 0.03 [micro]M and [V.sub.max] = 0.39 [+ or -] 0.02 nmol/min/mg of protein) and a glucose transporter with moderate affinity and capacity ([K.sub.m] = 46 [+ or -] 3 [micro]M and [V.sub.max] = 48 [+ or -] 1 nmol/min/mg of protein). The contribution of the trehalose transporter is important only in trehalose-poor environments (trehalose concentrations up to 6 [micro]M); at higher concentrations trehalose is assimilated primarily via trehalase and the glucose transport system. Trehalose uptake was constitutive, but the activity decreased 60% in response to osmotic stress. The nature of the trehalose transporter and the physiological relevance of these findings are discussed.

Plasmodium knowlesi poses a health threat throughout Southeast Asian communities and currently causes most cases of malaria in Malaysia. This zoonotic parasite species has been studied in Macaca mulatta (rhesus monkeys) as a model for severe malarial infections, chronicity, and antigenic variation. The phenomenon of Plasmodium antigenic variation was first recognized during rhesus monkey infections. Plasmodium -encoded variant proteins were first discovered in this species and found to be expressed at the surface of infected erythrocytes, and then named the Schizont-Infected Cell Agglutination (SICA) antigens. SICA expression was shown to be spleen dependent, as SICA expression is lost after P. knowlesi is passaged in splenectomized rhesus. Here we present data from longitudinal P. knowlesi infections in rhesus with the most comprehensive analysis to date of clinical parameters and infected red blood cell sequestration in the vasculature of tissues from 22 organs. Based on the histopathological analysis of 22 tissue types from 11 rhesus monkeys, we show a comparative distribution of parasitized erythrocytes and the degree of margination of the infected erythrocytes with the endothelium. Interestingly, there was a significantly higher burden of parasites in the gastrointestinal tissues, and extensive margination of the parasites along the endothelium, which may help explain gastrointestinal symptoms frequently reported by patients with P. knowlesi malarial infections. Moreover, this margination was not observed in splenectomized rhesus that were infected with parasites not expressing the SICA proteins. This work provides data that directly supports the view that a subpopulation of P. knowlesi parasites cytoadheres and sequesters, likely via SICA variant antigens acting as ligands. This process is akin to the cytoadhesive function of the related variant antigen proteins, namely Erythrocyte Membrane Protein-1, expressed by Plasmodium falciparum. [ABSTRACT FROM AUTHOR]

Background: Plasmodium vivax can cause severe malaria. The total parasite biomass during infections is correlated with the severity of disease but not necessarily quantified accurately by microscopy. This finding has raised the question whether there could be sub-populations of parasites that are not observed in peripheral blood smears but continue to contribute to the increase in parasite numbers that drive pathogenesis. Non-human primate infection models utilizing the closely related simian malaria parasite Plasmodium cynomolgi hold the potential for quantifying the magnitude of possibly unobserved infected red blood cell (iRBC) populations and determining how the presence of this hidden reservoir correlates with disease severity. Methods: Time series data tracking the longitudinal development of parasitaemia in five Macaca mulatta infected with P. cynomolgi were used to design a computational model quantifying iRBCs that circulate in the blood versus those that are not detectable and are termed here as 'concealed'. This terminology is proposed to distinguish such observations from the deep vascular and widespread 'sequestration' of Plasmodium falciparum iRBCs, which is governed by distinctly different molecular mechanisms. Results: The computational model presented here clearly demonstrates that the observed growth data of iRBC populations are not consistent with the known biology and blood-stage cycle of P. cynomolgi. However, the discrepancies can be resolved when a sub-population of concealed iRBCs is taken into account. The model suggests that the early growth of a hidden parasite sub-population has the potential to drive disease. As an alternative, the data could be explained by the sequential release of merozoites from the liver over a number of days, but this scenario seems less likely. Conclusions: Concealment of a non-circulating iRBC sub-population during P. cynomolgi infection of M. mulatta is an important aspect of this successful host-pathogen relationship. The data also support the likelihood that a subpopulation of iRBCs of P. vivax has a comparable means to become withdrawn from the peripheral circulation. This inference has implications for understanding vivax biology and pathogenesis and stresses the importance of considering a concealed parasite reservoir with regard to vivax epidemiology and the quantification and treatment of P. vivax infections. [ABSTRACT FROM AUTHOR]

We investigated by 13C NMR the turnover of the H3 deuterons of (2-13C) glutamate and (2-13C) glutamine in the brain of partially deuterated rats. Adult animals (150–200 g) fed ad libitum received 50%2H2O or tap water 9 days before infusing (1-13C) glucose or (2-13C) acetate for 5, 10, 15, 30, 60, or 90 min. The brains were then funnel-frozen and acid extracts were prepared and analyzed by high-resolution 13C NMR. The deuteration of one or the two H3 hydrogens of (2-13C) glutamate or glutamine resulted in single (−0.07 ppm) or double (−0.14 ppm) isotopic shifts upfield of the corresponding C2 perprotonated resonance, demonstrating two sequential deuteration steps. The faster monodeuteration generated 3 R or 3 S (2-13C, 3-2H) glutamate or glutamine through the alternate activities of cerebral aconitase or isocitrate dehydrogenase, respectively. The slower process produced bideuterated (2-13C, 3,3′-2H2) glutamate or glutamine through the consecutive activity of both enzymes. The kinetics of deuteration was fitted to a Michaelis–Menten model including the apparent Km′ and Vmax′ values for the observed deuterations. Our results revealed different kinetic constants for the alternate and consecutive deuterations, suggesting that these processes were caused by the different cytosolic or mitochondrial isoforms of aconitase and isocitrate dehydrogenase, respectively. The deuterations of (2-13C) glutamate or glutamine followed also different kinetics from (1-13C) glucose or (2-13C) acetate, revealing distinct deuteration environments in the neuronal or glial compartments. [ABSTRACT FROM AUTHOR]