Abstract: Atherosclerosis is the main cause of mortality and morbidity in the western world. Atherosclerosis is a chronic disease defined by life-long processes, with multiple actors playing a role at different biological and time scales. Patient-specific in silico models and simulations can help to understand better the mechanisms of atherosclerosis formation, potentially improving patient management. A conceptual and computational multiscale framework for the modelling of atherosclerosis formation at its early stage was created from the integration of a fluid mechanics model and a biochemical model. The fluid mechanics model describes the interaction between arterial endothelium and blood flow using an artery-specific approach. The low density lipoprotein (LDL) oxidation and consequent immune reaction leading to chronic inflammatory process at the basis of plaque formation was described in the biochemical model. The integration of these modelling approaches led to the creation of a computational framework, an effective tool for the modelling of atherosclerosis plaque development. The model presented in this study was able to capture key features of atherogenesis such as the location of pro-atherogenic areas and to reproduce the formation of plaques detectable from in vivo observations. This framework is being currently tested at University College Hospital (UCH).

Abstract: Tumor growth is a very challenging issue of capital importance to address therapy and patient management. Since it is difficult to follow the cancer natural history in humans, animal models are largely investigated. In particular, xeno-transplants are often performed on previously immune-depressed mice in order to get information about both macroscopic (growth rate) and microscopical (at cellular and genomic level) features. Previous studies showed that following multiple transplants tumors grow faster, and this fact was commonly assumed to prove the occurrence of mutations whose rate depended on the transplantation passage due to some sort of genetic instability. A recent paper reports data from a very interesting experiment, where two different clones are monitored through multipassage xeno-trasplant. We use these data in order to validate a two-population Gompertz model which assume a constant mutation rate but takes into account for the timing of multiple transplants.

Abstract: A multiscale model for an improved description of radial particle transport described by the density continuity equation in large-scale plasma edge simulations for tokamak fusion devices is presented. It includes the effects of mesoscale drift-fluid dynamics on the macroscale profiles and vice versa. The realization of the multiscale model in form of the coupled code system B2-ATTEMPT is outlined. A procedure employed to efficiently determine the averaged mesoscale terms using a nonparametric trend test, the Reverse Arrangements Test, is described. Results of stationary, self-consistent B2-ATTEMPT simulations are compared to prior simulations for experiments at the TEXTOR tokamak, making a first evaluation of the predicted magnitude of radial particle transport possible.

Abstract: A fusion plasma is a complex object involving a wide range of physics phenomena occurring at different scales. When building multiscale or multi-physics applications, an interesting approach (formalized in the Multiscale Modelling and Simulation Framework) consists in coupling single scale components (where a scale can be either spatial, temporal or refer to a different physics or numeric model), making each single component easier to develop, validate and maintain. Such coupling has been investigated within the EFDA ITM-TF task force, by using a common data structure to interface every single scale component, and a workflow manager to pilot the simulation. When such a workflow has to run in parallel, a possible approach consists in running the simulation platform within a regular parallel allocation in a single computer. The Integrated Plasma Simulator has been based on such a principle: it runs in a single (possibly very large) allocation and handles internally the dynamic scheduling of each component considering several layers of abstraction for the parallelism. We have implemented within the IPS platform two fusion workflows with different computational needs: an acyclic (loose-coupling) chain composed of high-resolution equilibrium reconstruction and MHD stability study, and a cyclic (tight-coupling) turbulence – transport time evolution. The acyclic case involves a parameter scan where the runtime of each single case can differ significantly, whereas the cyclic case is composed of codes which have to be executed in sequence with different level of parallelism and computational cost. This contribution presents briefly the characteristics of the IPS platform and compares them to other platforms used in the fusion community (Kepler, Muscle). Then implementation details are given about the wrappers, which are required to embed legacy codes coming from the ITM community into the IPS platform. Finally, targeted cyclic and acyclic workflows and their characteristics are presented as well as their performance in different configurations.

Abstract: A spectral method for kinetic plasma simulations based on the expansion of the velocity distribution function in a variable number of Hermite polynomials is presented. The method is based on a set of non-linear equations that is solved to determine the coefficients of the Hermite expansion satisfying the Vlasov and Poisson equations. In this paper, we first show that this technique combines the fluid and kinetic approaches into one framework. Second, we present an adaptive strategy to increase and decrease the number of Hermite functions dynamically during the simulation. The technique is applied to the Landau damping and two-stream instability test problems. Performance results show 21% and 47% saving of total simulation time in the Landau and two-stream instability test cases, respectively.