Instances Of Model Research Articles

Modeling the Subclonal Evolution of Cancer Cell Populations.

Increasing evidence shows that tumor clonal architectures are often the consequence of a complex branching process, yet little is known about the expected dynamics and extent to which these divergent subclonal expansions occur. Here, we develop and implement more than 88,000 instances of a stochastic evolutionary model simulating genetic drift and neoplastic progression. Under different combinations of population genetic parameter values, including those estimated for colorectal cancer and glioblastoma multiforme, the distribution of sizes of subclones carrying driver mutations had a heavy right tail at the time of tumor detection, with only 1 to 4 dominant clones present at ≥10% frequency. In contrast, the vast majority of subclones were present at <10% frequency, many of which had higher fitness than currently dominant clones. The number of dominant clones (≥10% frequency) in a tumor correlated strongly with the number of subclones (<10% of the tumor). Overall, these subclones were frequently below current standard detection thresholds, frequently harbored treatment-resistant mutations, and were more common in slow-growing tumors.Significance: The model presented in this paper addresses tumor heterogeneity by framing expectations for the number of resistant subclones in a tumor, with implications for future studies of the evolution of therapeutic resistance. Cancer Res; 78(3); 830-9. ©2017 AACR.

Experimentally integrated dynamic modelling for intuitive optimisation of cell based processes and manufacture

Dynamic mechanistic modelling of cell culture is a key tool in bioprocess development to support optimisation and risk assessment. However, the approach is underutilised in the bioprocess industry due to challenges including lack of accessible tools to support a structured approach, the difficulty of realising computationally tractable (low parameter) yet realistic models, and the specialised skill sets required. We have proposed that these issues could be partly addressed by developing a parsimonious framework comprising a set of model building blocks, based on the ordinary differential equation modelling paradigm, representing common cell culture dynamics and modulation thereof. The framework is designed to avoid obvious pathological behaviours. Further, specific model instances within the framework can be simply visualised as a directed graph with vertices representing system species, dynamics and modulations, and arcs representing the interactions between them. The directed graph can be extended to describe the timing and nature of experimental interventions. A visual and intuitive route to describing models with an associated mathematical framework enables realisation in a software interface and integration with standard mathematical tools such as those for sensitivity analysis and parameter estimation. Such a framework is sufficient to represent some of the simple mechanisms underpinning bioprocesses that nonetheless lead to highly non-linear and counterintuitive outcomes. It also has a relatively low learning burden for users without formal mathematical training. The concept could be extended to include, for example, partial differential equation-based approaches to stochastic or spatially complex systems built up from a small number of parametrically parsimonious and well-behaved building blocks.

Comparing Structural Identification Methodologies for Fatigue Life Prediction of a Highway Bridge

Accurate measurement-data interpretation leads to increased understanding of structural behaviour and enhanced asset-management decision making. In this paper, four data-interpretation methodologies, residual minimization, traditional Bayesian model updating, modified Bayesian model updating (with an $L_\infty$-norm-based Gaussian likelihood function) and error-domain model falsification (EDMF), a method that rejects models that have unlikely differences between predictions and measurements, are compared. In the modified Bayesian model updating methodology, a correction is used in the likelihood function to account for the effect of a finite number of measurements on posterior probability-density functions. The application of these data-interpretation methodologies for condition assessment and fatigue-life prediction is illustrated on a highway steel-concrete composite bridge having four spans with a total length of 219m. A detailed 3D finite-element plate and beam model of the bridge and weigh-in-motion data are used to obtain the time-stress response at a fatigue critical location along the bridge span. The time stress-response, presented as a histogram, is compared to measured strain responses either to update prior knowledge of model parameters using residual minimization and Bayesian methodologies or to obtain candidate model instances using the EDMF methodology. It is concluded that the EDMF and modified Bayesian model updating methodologies provide robust prediction of fatigue-life compared with residual minimization and traditional Bayesian model updating in the presence of correlated non-Gaussian uncertainty. EDMF has additional advantages due to ease of understanding and applicability for practising engineers, thus enabling incremental asset-management decision making over long service lives. Finally, parallel implementations of EDMF using grid sampling have lower computations times than implementations using adaptive sampling.

Shape Recognition Based on Projected Edges and Global Statistical Features

A combined shape descriptor for object recognition is presented, along with an offline and online learning method. The descriptor is composed of a local edge-based part and global statistical features. We also propose a two-level, nearest neighborhood type multiclass classification method, in which classes are bounded, defining an inherent rejection region. In the first stage, global features are used to filter model instances, in contrast to the second stage, in which the projected edge-based features are compared. Our experimental results show that the combination of independent features leads to increased recognition robustness and speed. The core algorithms map easily to cellular architectures or dedicated VLSI hardware.

Simulating Head Losses in Hydraulic Drop Structures in SWMM, with a Little Help from CFD

The City of Portland, Oregon has cause to model instances of street flooding in the vicinity of hydraulic drop structures that feed into the Willamette combine…

Model Uncertainty Analysis for Metabolic Network Inference: A Case Study in Bayesian Model Averaging

A distinguishing feature of systems biology is the interrogation of models as a means of making predictions or generating deeper understanding of the systems under study. However, when using a given data set to address a specific question, a unique and provably correct model formulation to apply is rarely known. Instead, a large selection of alternative formulations of varying scopes ensues from combinatorial composition of entities. In this scenario, computational methods that allow us to make statistically valid inferences and predictions, while accounting for the uncertainty in model formulation are desired.We investigate into Bayesian Model Averaging (BMA), which accounts for model uncertainty by considering an ensemble of candidate models instead of a single model instance. To show the computational tractability of BMA, we perform model uncertainty analysis for a realistically sized reaction network from the domain of metabolic flux analysis, featuring an ensemble of millions of models. This is made possible using a Markov Chain Monte Carlo (MCMC) method, tailored to handle parameter and model structure uncertainty simultaneously. To investigate the computational burden of solving the multi-model problem, a super-model is created that includes the reactions of all models in the multi-model problem. The computational burden of the multi-model problem is compared to that of conventional MCMC inference on the single super-model. The comparison yields the surprising insight that the multi-model problem is computationally less expensive than the single super-model problem. Furthermore, we demonstrate, with the example at hand, that BMA yields valid structural network inferences.

Multiscaled Simulation Methodology for Neuro-Inspired Circuits Demonstrated with an Organic Memristor

Organic memristors are promising molecular electronic devices for neuro-inspired on-chip learning applications. In this paper, we present a numerically efficient compact model suitable for $Fe(bpy)_3^{2+}$ organic memristors operating according to an intramolecular charge transfer switching mechanism. This compact model, being physics-based and relying on electrical characterizations and parametric extractions performed on test structures, is especially efficient in pulsed mode and describes the conductance variations for both SET and RESET regimes. Using this model, a dynamic multiscale simulation approach has been set-up to extend the model from individual devices to larger model systems that learn progressively through time. To verify the soundness and highlight emergent properties of the organic memristors, instances of the compact model have been simulated within a simple neuromorphic design that co-integrates with CMOS neurons. In addition, a larger supervised learning system using the new compact model is demonstrated. These successful tests suggest our model might be of interest to neuromorphic designers.

BrainFrame: a node-level heterogeneous accelerator platform for neuron simulations

Objective. The advent of high-performance computing (HPC) in recent years has led to its increasing use in brain studies through computational models. The scale and complexity of such models are constantly increasing, leading to challenging computational requirements. Even though modern HPC platforms can often deal with such challenges, the vast diversity of the modeling field does not permit for a homogeneous acceleration platform to effectively address the complete array of modeling requirements. Approach. In this paper we propose and build BrainFrame, a heterogeneous acceleration platform that incorporates three distinct acceleration technologies, an Intel Xeon-Phi CPU, a NVidia GP-GPU and a Maxeler Dataflow Engine. The PyNN software framework is also integrated into the platform. As a challenging proof of concept, we analyze the performance of BrainFrame on different experiment instances of a state-of-the-art neuron model, representing the inferior-olivary nucleus using a biophysically-meaningful, extended Hodgkin–Huxley representation. The model instances take into account not only the neuronal-network dimensions but also different network-connectivity densities, which can drastically affect the workload’s performance characteristics. Main results. The combined use of different HPC technologies demonstrates that BrainFrame is better able to cope with the modeling diversity encountered in realistic experiments while at the same time running on significantly lower energy budgets. Our performance analysis clearly shows that the model directly affects performance and all three technologies are required to cope with all the model use cases. Significance. The BrainFrame framework is designed to transparently configure and select the appropriate back-end accelerator technology for use per simulation run. The PyNN integration provides a familiar bridge to the vast number of models already available. Additionally, it gives a clear roadmap for extending the platform support beyond the proof of concept, with improved usability and directly useful features to the computational-neuroscience community, paving the way for wider adoption.

Prediction of Nonlinear Distortion in Wideband Active Antenna Arrays

In this paper, we propose a technique for comprehensive analysis of nonlinear and dynamic characteristics of multi-antenna transmitters (TXs). The analysis technique is enabled by the development of a Volterra series-based dual-input model for power amplifiers (PAs), which is capable of considering the joint effects of PA nonlinearity, antenna crosstalk, and mismatch for wideband modulated signals. By combining multiple instances of the PA model with linear dynamic antenna simulations, we develop the analysis technique. The proposed method allows the prediction of the output signal of every antenna in an arbitrarily sized TX array, as well as the total far-field radiated wave of the TX for any input signal with low computational effort. A 2.12-GHz four-element TX demonstrator based on GaAs PAs is implemented to verify simulation results with measurements. The proposed technique is a powerful tool to study hardware characteristics as, for example, the effects of antenna design and element spacing. It can be used in cases where experiments are not feasible, and thus aid the development of next generation wireless systems.

Poster Presentations

Application of computer simulation for predicting cancer progression/remission/ recurrence is still underestimated by clinicians. This is mainly due to the lack of tumor modeling approaches, which are both reliable and realistic computationally. We present here the concept of a viable prediction/correction system for predicting tumor dynamics. It is very similar, in spirit, to that used in weather forecast and climate modeling. The system is based on the supermodeling technique where the supermodel consists of a few coupled instances (sub-models) of a generic coarse-grained tumor model. Consequently, the latent and fine-grained cancer properties, not included in the generic model, e.g., reflecting microscopic phenomena and other unpredictable events influencing tumor dynamics, are hidden in sub-models coupling parameters, which can be learned from incoming real data. Thus instead of matching hundreds of parameters for multi-scale tumor models, we need to fit only several values of coupling coefficients between sub-models to simulate the current tumor status.

Sparse coding based classifier ensembles in supervised and active learning scenarios for data classification

Sparse coding and dictionary learning has recently gained great interest in signal, image and audio processing applications through representing each problem instance by a sparse set of atoms. This also allows us to obtain different representations of feature sets in machine learning problems. Thus, different feature views for classifier ensembles can be obtained using sparse coding. On the other hand, nowadays unlabelled data is abundant and active learning methods with single and classifier ensembles received great interest. In this study, Random Subspace Dictionary Learning (RDL) and Bagging Dictionary Learning (BDL) algorithms are examined by learning ensembles of dictionaries through feature/instance subspaces. Besides, ensembles of dictionaries are evaluated under active learning framework as promising models and they are named as Active Random Subspace Dictionary Learning (ARDL) and Active Bagging Dictionary Learning (ABDL) algorithms. Active learning methods are compared with their Support Vector Machines counterparts. The experiments on eleven datasets from UCI and OpenML repositories has shown that selecting instance and feature subspaces for dictionary learning model increases the number of correctly classified instances for the most of the data sets while SVM has superiority over all of the applied models. Furthermore, using an active learner generally increases the chance of improved classification performance as the number of iterations is increased.

Globally optimized packings of non-uniform size spheres in $$\mathbb {R}^{d}$$ R d : a computational study

In this work we discuss the following general packing problem: given a finite collection of d-dimensional spheres with (in principle) arbitrarily chosen radii, find the smallest sphere in $$\mathbb {R}^{d}$$ that contains the given d-spheres in a non-overlapping arrangement. Analytical (closed-form) solutions cannot be expected for this very general problem-type: therefore we propose a suitable combination of constrained nonlinear optimization methodology with specifically designed heuristic search strategies, in order to find high-quality numerical solutions in an efficient manner. We present optimized sphere configurations with up to $$n = 50$$ spheres in dimensions $$d = 2, 3, 4, 5$$ . Our numerical results are on average within 1% of the entire set of best known results for a well-studied model-instance in $$\mathbb {R}^{2}$$ , with new (conjectured) packings for previously unexplored generalizations of the same model-class in $$\mathbb {R}^{d}$$ with $$d= 3, 4, 5.$$ Our results also enable the estimation of the optimized container sphere radii and of the packing fraction as functions of the model instance parameters n and 1 / n, respectively. These findings provide a general framework to define challenging packing problem-classes with conjectured numerical solution estimates.

Solution to urn models of pairwise interaction with application to social, physical, and biological sciences.

We investigate a family of urn models that correspond to one-dimensional random walks with quadratic transition probabilities that have highly diverse applications. Well-known instances of these two-urn models are the Ehrenfest model of molecular diffusion, the voter model of social influence, and the Moran model of population genetics. We also provide a generating function method for diagonalizing the corresponding transition matrix that is valid if and only if the underlying mean density satisfies a linear differential equation and express the eigenvector components as terms of ordinary hypergeometric functions. The nature of the models lead to a natural extension to interaction between agents in a general network topology. We analyze the dynamics on uncorrelated heterogeneous degree sequence networks and relate the convergence times to the moments of the degree sequences for various pairwise interaction mechanisms.

Extended Features of Task Models for Specifying Cooperative Activities

In this paper, an extended version of task models is discussed that allows detailed specifications of cooperative activities. Within the presented specification CoTaL (Cooperative Task Language), there exist two complementary types of task models called role model and team model. One or more instances of each role model describe the specific activities of actors. The team model represents joined activities and reflects progression in cooperation between role instances. For each scenario there exists one instance of the team model. Preconditions and events can be assigned to tasks and refer to one or all running instances of a role model. An event can be a starting or finishing trigger and is activated as result of task executions. Additionally, variables can be defined. They are bound within the specified context during runtime and get the value of the identifier of a certain role instance. In this way, communication and collaboration between different actors can be specified. Tasks of a team model cannot be performed directly but present the result of the execution of other (role) models only. However, a team model can restrict the execution of role model instances. It is shown how such models can be used to specify the activities in a smart meeting room. Snapshots of their simulation in CoTaSE (Cooperative Task Specification Environment) are presented. Additionally to the local implementation there exists an implementation in a cloud. It allows real cooperative executions of tasks.

Structure-based control of complex networks with nonlinear dynamics

What can we learn about controlling a system solely from its underlying network structure? Here we adapt a recently developed framework for control of networks governed by a broad class of nonlinear dynamics that includes the major dynamic models of biological, technological, and social processes. This feedback-based framework provides realizable node overrides that steer a system toward any of its natural long-term dynamic behaviors, regardless of the specific functional forms and system parameters. We use this framework on several real networks, identify the topological characteristics that underlie the predicted node overrides, and compare its predictions to those of structural controllability in control theory. Finally, we demonstrate this framework's applicability in dynamic models of gene regulatory networks and identify nodes whose override is necessary for control in the general case but not in specific model instances.

Stochastic fleet deployment models for public bicycle rental systems

ABSTRACTThis paper presents two stochastic bike deployment (SBD) models that determine the optimal number of bicycles allocated to each station in a leisure-oriented public bicycle rental system with stochastic demands. The SBD models represent the stochastic demands using a set of scenarios with given probabilities. A multilayer bike-flow time-space network is constructed for developing the models, where each layer corresponds to a given demand scenario and effectively describes bicycle flows in the spatial and temporal dimensions. As a result, the models are formulated as the integer multi-commodity network flow problem, which is characterized as NP-hard. We propose a heuristic to efficiently obtain good quality solutions for large-size model instances. Test instances are generated using real data from a bicycle rental system in Taiwan to evaluate the performance of the models and the solution algorithm. The test results show that the models can help the system operator of a public bicycle system make effective fleet deployment decisions.

An evolutionary approach to a combined mixed integer programming model of seaside operations as arise in container ports

This paper puts forward an integrated optimisation model that combines three distinct problems, namely berth allocation, quay crane assignment, and quay crane scheduling that arise in container ports. Each one of these problems is difficult to solve in its own right. However, solving them individually leads almost surely to sub-optimal solutions. Hence, it is desirable to solve them in a combined form. The model is of the mixed-integer programming type with the objective being to minimize the tardiness of vessels and reduce the cost of berthing. Experimental results show that relatively small instances of the proposed model can be solved exactly using CPLEX. Large scale instances, however, can only be solved in reasonable times using heuristics. Here, an implementation of the genetic algorithm is considered. The effectiveness of this implementation is tested against CPLEX on small to medium size instances of the combined model. Larger size instances were also solved with the genetic algorithm, showing that this approach is capable of finding the optimal or near optimal solutions in realistic times.

Multiple Authority Delegation in Art Authentication

In this paper, I expand upon the research on authority delegation begun by Overgaard and myself in our 2016 paper Authority Delegation. I argue that singular authority delegation – in which a community delegates authority over a given topic to a single expert community – should be distinguished from cases of multiple authority delegation. A community engages in multiple authority delegation iff that community delegates authority over a given topic to more than one expert community. Furthermore, multiple authority delegation can be further divided into two types: hierarchical and non-hierarchical. I examine two cases of authority delegation in the art market and argue that these cases model instances of hierarchical authority delegation.Suggested Modifications[Sciento-2017-0007]: Accept the following definitions of subtypes of authority delegation: Singular authority delegation ≡ community A is said to engage in a relationship of singular authority delegation over topic x iff community A delegates authority over topic x to exactly one community.Multiple authority delegation ≡ community A is said to engage in a relationship of multiple authority delegation over topic x iff community A delegates authority over topic x to more than one community. Hierarchical authority delegation ≡ a sub-type of multiple authority delegation where different communities are delegated different degrees of authority over topic x.Non-hierarchical authority delegation ≡ a sub-type of multiple authority delegation where different communities are delegated the same degree of authority over topic x. [Sciento-2017-0008]: Accept the following reconstruction of the contemporary authority delegation structure in the art market regarding the works of Monet: A work claimed to be by Monet is authentic if it is considered authentic by the Wildenstein Institute.[Sciento-2017-0009]: Accept the following reconstruction of the contemporary authority delegation structure in the art market regarding the works of Picasso: A work claimed to be by Picasso is authentic if it is has been certified as authentic by both Maya Widmaier-Picasso and Claude Ruiz-Picasso.[Sciento-2017-0010]: Accept the following reconstruction of the authority delegation structure in the art market regarding the works of Modigliani between 1997 and 2015: A work claimed to be by Modigliani is authentic iff (1) it is in the Ceroni catalogue raisonné or (2) if it is not in catalogue and has been certified as authentic by Marc Restellini.[Sciento-2017-0011]: Accept the following reconstruction of the contemporary authority delegation structure in the art market regarding the works of Renoir: A work claimed to be by Renoir is authentic iff (1) it has been certified as authentic by the Wildenstein institute or (2) it has not been dismissed by the Wildenstein institute and it is included in the Bernheim-Jeune catalogue.

Hybrid annealing: Coupling a quantum simulator to a classical computer

Finding the global minimum in a rugged potential landscape is a computationally hard task, often equivalent to relevant optimization problems. Simulated annealing is a computational technique which explores the configuration space by mimicking thermal noise. By slow cooling, it freezes the system in a low-energy configuration, but the algorithm often gets stuck in local minima. In quantum annealing, the thermal noise is replaced by controllable quantum fluctuations, and the technique can be implemented in modern quantum simulators. However, quantum-adiabatic schemes become prohibitively slow in the presence of quasidegeneracies. Here we propose a strategy which combines ideas from simulated annealing and quantum annealing. In such hybrid algorithm, the outcome of a quantum simulator is processed on a classical device. While the quantum simulator explores the configuration space by repeatedly applying quantum fluctuations and performing projective measurements, the classical computer evaluates each configuration and enforces a lowering of the energy. We have simulated this algorithm for small instances of the random energy model, showing that it potentially outperforms both simulated thermal annealing and adiabatic quantum annealing. It becomes most efficient for problems involving many quasi-degenerate ground states.

Beyond-quantum modeling of question order effects and response replicability in psychological measurements

A general tension-reduction (GTR) model was recently considered (Aerts and Sassoli de Bianchi 2015a, 2015b) to derive quantum probabilities as (universal) averages over all possible forms of non-uniform fluctuations and explain their considerable success in describing experimental situations also outside of the domain of physics, for instance in quantum models of cognition and decision. This result also highlighted the possibility of observing violations of the predictions of the Born rule in those situations where the averaging would not be large enough, or would be altered because of the combination of multiple measurements. We show that this is indeed the case in typical psychological measurements exhibiting question order effects, as the data require the larger (non-Hilbertian) framework of the GTR-model to receive an exact mathematical description. We also consider an unsolved problem of quantum cognition: response replicability. It has been noted that if question order effects and response replicability would occur together, the situation could not be handled anymore by quantum theory. Yet, we show that it can be easily and naturally described in the GTR-model. Based on these findings, we motivate the adoption in cognitive science of a hidden-measurements interpretation of the quantum formalism, and of its GTR-model generalization, as a natural interpretational framework explaining the data of psychological measurements on conceptual entities.