VECMAtk: a scalable verification, validation and uncertainty quantification toolkit for scientific simulations.

We present the VECMA toolkit (VECMAtk), a flexible software environment for single and multiscale simulations that introduces directly applicable and reusable procedures for verification, validation (V&V), sensitivity analysis (SA) and uncertainty quantication (UQ). It enables users to verify key aspects of their applications, systematically compare and validate the simulation outputs against observational or benchmark data, and run simulations conveniently on any platform from the desktop to current multi-petascale computers. In this sequel to our paper on VECMAtk which we presented last year [1] we focus on a range of functional and performance improvements that we have introduced, cover newly introduced components, and applications examples from seven different domains such as conflict modelling and environmental sciences. We also present several implemented patterns for UQ/SA and V&V, and guide the reader through one example concerning COVID-19 modelling in detail. This article is part of the theme issue 'Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification in silico'.

Sensitivity-driven simulation development: a case study in forced migration.

This paper presents an approach named sensitivity-driven simulation development (SDSD), where the use of sensitivity analysis (SA) guides the focus of further simulation development and refinement efforts, avoiding direct calibration to validation data. SA identifies assumptions that are particularly pivotal to the validation result, and in response model ruleset refinement resolves those assumptions in greater detail, balancing the sensitivity more evenly across the different assumptions and parameters. We implement and demonstrate our approach to refine agent-based models of forcibly displaced people in neighbouring countries. Over 70.8 million people are forcibly displaced worldwide, of which 26 million are refugees fleeing from armed conflicts, violence, natural disaster or famine. Predicting forced migration movements is important today, as it can help governments and NGOs to effectively assist vulnerable migrants and efficiently allocate humanitarian resources. We use an initial SA iteration to steer the simulation development process and identify several pivotal parameters. We then show that we are able to reduce the relative sensitivity of these parameters in a secondary SA iteration by approximately 54% on average. This article is part of the theme issue 'Reliability and reproducibility in computational science: implementing verification, validation and uncertainty quantification in silico'.

Stratum corneum lipid matrix: Location of acyl ceramide and cholesterol in the unit cell of the long periodicity phase.

The extracellular lipid matrix in the skin's outermost layer, the stratum corneum, is crucial for the skin barrier. The matrix is composed of ceramides (CERs), cholesterol (CHOL) and free fatty acids (FFAs) and involves two lamellar phases: the short periodicity phase (SPP) and the long periodicity phase (LPP). To understand the skin barrier thoroughly, information about the molecular arrangement in the unit cell of these lamellar phases is paramount. Previously we examined the molecular arrangement in the unit cell of the SPP. Furthermore X-ray and neutron diffraction revealed a trilayer arrangement of lipids within the unit cell of the LPP [D. Groen et al., Biophysical Journal, 97, 2242-2249, 2009]. In the present study, we used neutron diffraction to obtain more details about the location of lipid (sub)classes in the unit cell of the LPP. The diffraction pattern revealed at least 8 diffraction orders of the LPP with a repeating unit of 129.6±0.5Å. To determine the location of lipid sub(classes) in the unit cell, samples were examined with either only protiated lipids or selectively deuterated lipids. The diffraction data obtained by means of D2O/H2O contrast variation together with a gradual replacement of one particular CER, the acyl CER, by its partly deuterated counterpart, were used to construct the scattering length density profiles. The acyl chain of the acyl CER subclass is located at a position of ~21.4±0.2Å from the unit cell centre of the LPP. The position and orientation of CHOL in the LPP unit cell were determined using tail and head-group deuterated forms of the sterol. CHOL is located with its head-group positioned ~26±0.2Å from the unit cell centre. This allows the formation of a hydrogen bond with the ester group of the acyl CER located in close proximity. Based on the positions of the deuterated moieties of the acyl CER, CHOL and the previously determined location of two other lipid subclasses [E.H. Mojumdar et al., Biophysical Journal, 108, 2670-2679, 2015], a molecular model is proposed for the unit cell of the LPP. In this model CHOL is located in the two outer layers of the LPP, while CER EOS is linking the two outer layers with the central lipid layers. Finally the two other lipid subclasses are predominantly located in the central layer of the LPP.

Performance of distributed multiscale simulations.

Multiscale simulations model phenomena across natural scales using monolithic or component-based code, running on local or distributed resources. In this work, we investigate the performance of distributed multiscale computing of component-based models, guided by six multiscale applications with different characteristics and from several disciplines. Three modes of distributed multiscale computing are identified: supplementing local dependencies with large-scale resources, load distribution over multiple resources, and load balancing of small- and large-scale resources. We find that the first mode has the apparent benefit of increasing simulation speed, and the second mode can increase simulation speed if local resources are limited. Depending on resource reservation and model coupling topology, the third mode may result in a reduction of resource consumption.

Localization of cholesterol and fatty acid in a model lipid membrane: a neutron diffraction approach.

The intercellular lipid matrix of the skin's stratum corneum serves to protect the body against desiccation and simultaneously limits the passage of drugs and other xenobiotics into the body. The matrix is made up of ceramides, free fatty acids, and cholesterol, which are organized as two coexisting crystalline lamellar phases. In studies reported here, we sought to use the technique of neutron diffraction, together with the device of isotopic (H/D) substitution, to determine the molecular architecture of the lamellar phase having a repeat distance of 53.9 ± 0.3 Å. Using hydrogenous samples as well as samples incorporating perdeuterated (C24:0) fatty acids and selectively deuterated cholesterol, the diffraction data obtained were used to construct neutron scattering length density profiles. By this means, the locations within the unit cell were determined for the cholesterol and fatty acids. The cholesterol headgroup was found to lie slightly inward from the unit cell boundary and the tail of the molecule located 6.2 ± 0.2 Å from the unit cell center. The fatty acid headgroups were located at the unit cell boundary with their acyl chains straddling the unit cell center. Based on these results, a molecular model is proposed for the arrangement of the lipids within the unit cell.

Flexible composition and execution of high performance, high fidelity multiscale biomedical simulations.

Multiscale simulations are essential in the biomedical domain to accurately model human physiology. We present a modular approach for designing, constructing and executing multiscale simulations on a wide range of resources, from laptops to petascale supercomputers, including combinations of these. Our work features two multiscale applications, in-stent restenosis and cerebrovascular bloodflow, which combine multiple existing single-scale applications to create a multiscale simulation. These applications can be efficiently coupled, deployed and executed on computers up to the largest (peta) scale, incurring a coupling overhead of 1-10% of the total execution time.

Disposition of ceramide in model lipid membranes determined by neutron diffraction.

The lipid matrix present in the uppermost layer of the skin, the stratum corneum, plays a crucial role in the skin barrier function. The lipids are organized into two lamellar phases. To gain more insight into the molecular organization of one of these lamellar phases, we performed neutron diffraction studies. In the diffraction pattern, five diffraction orders were observed attributed to a lamellar phase with a repeat distance of 5.4 nm. Using contrast variation, the scattering length density profile could be calculated showing a typical bilayer arrangement. To obtain information on the arrangement of ceramides in the unit cell, a mixture that included a partly deuterated ceramide was also examined. The scattering length density profile of the 5.4-nm phase containing this deuterated ceramide demonstrated a symmetric arrangement of the ceramides with interdigitating acyl chains in the center of the unit cell.

New insights into the stratum corneum lipid organization by X-ray diffraction analysis.

The characteristic 13-nm lamellar phase that is formed by lipids in the outermost layer of the skin, the stratum corneum (SC), is very important for the barrier function of the skin. To gain more insight into the molecular organization of this lamellar phase, we performed small-angle x-ray diffraction (SAXD) using various lipid mixtures mimicking the lipid composition in SC. In the SAXD pattern of each mixture, at least seven diffraction orders were observed, attributed to the lamellar phase with a repeat distance ranging from 12.1 to 13.8 nm. Using the sampling method based on the variation in repeat distance, we selected phase angles for the first six diffraction orders. Using these phase angles for the lamellar phase, a high-resolution electron density distribution could be calculated. Subsequently, from SAXD patterns of isolated SC, the electron density distribution of the lamellar phase was also calculated and appeared to be very similar to that in the lipid mixtures. This demonstrates that the lipid mixtures serve as an excellent model for the lipid organization in SC, not only with respect to the repeat distance, but also in terms of the electron density distribution within the unit cell.

Interference of polychlorinated biphenyls in hepatic and brain thyroid hormone metabolism in fetal and neonatal rats.

The effects of prenatal oral administration of 0.2, 0.6, and 1.8 mg/kg body wt of 3,3',4,4'5,5'-hexachlorobiphenyl (HCB) on Day 1 of gestation and a combination of 1 mg/kg 3,3',4,4'-tetrachlorobiphenyl (TCB) from Day 2 to Day 18 with 0.6 mg HCB/kg body wt on Day 1 of gestation on thyroid hormone status and peripheral thyroid metabolism were studied in pregnant Wistar rats and their fetuses and offspring. Plasma total thyroxine and free thyroxine levels were reduced by HCB in a dose-dependent fashion in pregnant rats (Days 12 and 20 of gestation) and neonates (Day 21 postpartum), while only a combined dose of HCB and TCB was effective in decreasing fetal thyroid hormone levels by 65% on Day 20 of gestation. The activity of type II thyroxine 5'-deiodinase (5'D-II), the enzyme responsible for the deiodination of thyroxine (T4) to biologically active triiodothyronine in the brain, was examined in whole brain homogenates in fetuses and neonates. Decreases in plasma thyroid hormones were accompanied by significant increases, up to 100%, in 5'D-II activity in brain homogenates from fetuses (Day 20 of gestation) and neonates (Days 7 and 21 postpartum). The glucuronidation of 125I-T4 by hepatic microsomes was increased by at least 100% relative to control levels by all treatments in fetuses (Day 20 of gestation) and increased at least 40% in neonates (Days 7 and 21 postpartum) by a dose of 0.6 and 1.8 mg HCB/kg and the combined dose. These data indicate that prenatal HCB and/or TCB administration result in increased peripheral T4 metabolism. The increase in 5'D-II activity suggests that local hypothyroidism occurs in the brains of fetal and neonatal rats exposed to HCB and/or TCB. Since these effects occur during a period in which thyroid hormones play an important role in brain maturation, they may help explain the mechanism of developmental neurotoxicity induced by polychlorinated biphenyls.

Countercurrent multistage fluidized bed reactor for immobilized biocatalysts: I. Modeling and simulation.

For the application of immobilized enzymes, fixed bed reactors are used almost exclusively. Fixed bed reactors have specific disadvantages, especially for processes with a deactivating catalyst. Therefore, we have studied a novel reactor type with continuous transport of the immobilized biocatalyst. Flow of biocatalyst is countercurrent to the substrate solution. Because of a stagewise reactor design, back-mixing of biocatalyst is very limited and transport is nearly plug flow. The reactor operates at a constant flow rate and conversion, due to constant holdup of catalytic activity. The reactor performance is compared with a configuration of fixed bed reactors. For reactions in the first-order regime, enzyme requirements in this new reactor are slightly less than for fixed bed processes. The multistage fluidized bed appears to be an attractive reactor design to use biocatalyst to a low residual activity. However, nonuniformity of the particles might affect plug flow transport of the biocatalyst. A laboratory scale reactor and experiments are described in Part II(1) of this series. Hydrodynamic design aspects of a multistage fluidized bed are discussed in more detail in Part III.(2).

Countercurrent multistage fluidized bed reactor for immobilized biocatalysts: II. Operation of a laboratory-scale reactor.

In Part I of this series,(1) we derived a model and made simulations for a multistage fluidized bed reactor (MFBR). It was concluded that the MFBR can be an attractive alternative for a fixed bed reactor when operated with a deactivating biocatalyst. In Part II of this series, the design of a laboratory-scale MFBR and its evaluation to investigate the practical feasibility of this reactor type, will be described. Experiments with a duration as long as 10 days were carried out successfully using immobilized glucose isomerase as a model reaction system. The results predicted by the model are in good agreement with the measured glucose concentration and biocatalyst activity gradients, indicating perfect mixing of the particles in the reactor compartments.The diameters of the biocatalyst particles used in the experiments showed a large spread, with the largest being 1.7 times the smallest. Therefore, an additional check was carried out, to make sure that the particles were not segregating according to size. Particles withdrawn from the reactor compartments were investigated using an image analyzer. Histograms of particle size distribution do not indicate segregation and it is concluded that the particles used have been mixed completely within the compartments. As a result, transport of biocatalyst is nearly plug flow.