Salem Fawaz Adra

Development of a three dimensional multiscale computational model of the human epidermis.

Transforming Growth Factor (TGF-β1) is a member of the TGF-beta superfamily ligand-receptor network. and plays a crucial role in tissue regeneration. The extensive in vitro and in vivo experimental literature describing its actions nevertheless describe an apparent paradox in that during re-epithelialisation it acts as proliferation inhibitor for keratinocytes. The majority of biological models focus on certain aspects of TGF-β1 behaviour and no one model provides a comprehensive story of this regulatory factors action. Accordingly our aim was to develop a computational model to act as a complementary approach to improve our understanding of TGF-β1. In our previous study, an agent-based model of keratinocyte colony formation in 2D culture was developed. In this study this model was extensively developed into a three dimensional multiscale model of the human epidermis which is comprised of three interacting and integrated layers: (1) an agent-based model which captures the biological rules governing the cells in the human epidermis at the cellular level and includes the rules for injury induced emergent behaviours, (2) a COmplex PAthway SImulator (COPASI) model which simulates the expression and signalling of TGF-β1 at the sub-cellular level and (3) a mechanical layer embodied by a numerical physical solver responsible for resolving the forces exerted between cells at the multi-cellular level. The integrated model was initially validated by using it to grow a piece of virtual epidermis in 3D and comparing the in virtuo simulations of keratinocyte behaviour and of TGF-β1 signalling with the extensive research literature describing this key regulatory protein. This research reinforces the idea that computational modelling can be an effective additional tool to aid our understanding of complex systems. In the accompanying paper the model is used to explore hypotheses of the functions of TGF-β1 at the cellular and subcellular level on different keratinocyte populations during epidermal wound healing.

Convergence Acceleration Operator for Multiobjective Optimization

A convergence acceleration operator (CAO) is described which enhances the search capability and the speed of convergence of the host multiobjective optimization algorithm. The operator acts directly in the objective space to suggest improvements to solutions obtained by a multiobjective evolutionary algorithm (MOEA). The suggested improved objective vectors are then mapped into the decision variable space and tested. This method improves upon prior work in a number of important respects, such as mapping technique and solution improvement. Further, the paper discusses implications for many-objective problems and studies the impact of the use of the CAO as the number of objectives increases. The CAO is incorporated with two leading MOEAs, the non-dominated sorting genetic algorithm and the strength Pareto evolutionary algorithm and tested. Results show that the hybridized algorithms consistently improve the speed of convergence of the original algorithm while maintaining the desired distribution of solutions. It is shown that the operator is a transferable component that can be hybridized with any MOEA.

Exploring Hypotheses of the Actions of TGF-β1 in Epidermal Wound Healing Using a 3D Computational Multiscale Model of the Human Epidermis

In vivo and in vitro studies give a paradoxical picture of the actions of the key regulatory factor TGF-β1 in epidermal wound healing with it stimulating migration of keratinocytes but also inhibiting their proliferation. To try to reconcile these into an easily visualized 3D model of wound healing amenable for experimentation by cell biologists, a multiscale model of the formation of a 3D skin epithelium was established with TGF-β1 literature–derived rule sets and equations embedded within it. At the cellular level, an agent-based bottom-up model that focuses on individual interacting units (keratinocytes) was used. This was based on literature-derived rules governing keratinocyte behavior and keratinocyte/ECM interactions. The selection of these rule sets is described in detail in this paper. The agent-based model was then linked with a subcellular model of TGF-β1 production and its action on keratinocytes simulated with a complex pathway simulator. This multiscale model can be run at a cellular level only or at a combined cellular/subcellular level. It was then initially challenged (by wounding) to investigate the behavior of keratinocytes in wound healing at the cellular level. To investigate the possible actions of TGF-β1, several hypotheses were then explored by deliberately manipulating some of these rule sets at subcellular levels. This exercise readily eliminated some hypotheses and identified a sequence of spatial-temporal actions of TGF-β1 for normal successful wound healing in an easy-to-follow 3D model. We suggest this multiscale model offers a valuable, easy-to-visualize aid to our understanding of the actions of this key regulator in wound healing, and provides a model that can now be used to explore pathologies of wound healing.

A comparative study of progressive preference articulation techniques for multiobjective optimisation

Multiobjective optimisation has traditionally focused on problems consisting of 2 or 3 objectives. Real-world problems often require the optimisation of a larger number of objectives. Research has shown that conclusions drawn from experimentations carried out on 2 or 3 objectives cannot be generalized for a higher number of objectives. The curse of dimensionality is a problem that faces decision makers when confronted with many objectives. Preference articulation techniques, and especially progressive preference articulation (PPA) techniques are effective methods for supporting the decision maker. In this paper, some of the most recent and most established PPA techniques are examined, and their utility for tackling many-objective optimisation problems is discussed and compared from the viewpoint of the decision maker.

A Diversity Management Operator for Evolutionary Many-Objective Optimisation

The proximity of an approximation set to the Pareto-optimal front of a multiobjective optimisation problem and the diversity of the solutions within the approximation set are two essential requirements in evolutionary multiobjective optimisation. These two requirements may be found to be in conflict with each other in many -objective optimisation scenarios deploying Pareto-dominance selection alongside active diversity promotion mechanisms. This conflict is hindering the optimisation process of some of the most established MOEAs and introducing problems such as the problem of dominance resistance and speciation. In this study, a diversity management operator (DMO) for controlling and promoting the diversity requirement in many -objective optimisation scenarios is introduced and tested on a set of test functions with increasing numbers (6 to 12) of objectives. The results achieved by the proposed strategy outperform results achieved by a reputed and representative MOEA in terms of both criteria: convergence and diversity.

A hybrid multi-objective evolutionary algorithm using an inverse neural network for aircraft control system design

This study introduces a hybrid multi-objective evolutionary algorithm (MOEA) for the optimization of aircraft control system design. The strategy suggested is composed mainly of two stages. The first stage consists of training an artificial neural network (ANN) with objective values as inputs and decision variables as outputs to model an approximation of the inverse of the objective function used. The second stage consists of a local improvement phase in objective space preserving objectives relationships, and a mapping process to decision variables using the trained ANN. Both the hybrid MOEA and the original MOEA were applied to an aircraft control system design application for assessment

Mutation Operators for Agent-Based Models

This short paper argues that agent-based models are an independent class of software application with their own unique properties, with the consequential need for the definition of suitable, tailored mutation operators. Testing agent-based models can be very challenging, and no established testing technique has yet been introduced for such systems. This paper discusses the application of mutation testing techniques, and mutation operators are proposed that can imitate potential programmer errors and result in faulty simulation runs of a model.

Hybrid multiobjective genetic algorithm with a new adaptive local search process

This paper is concerned with a specific brand of evolutionary algorithms: Memetic algorithms. A new local search technique with an adaptive neighborhood setting process is introduced and assessed against a set of test functions presenting different challenges. Two performance criteria were assessed: the convergence of the achieved results towards the true Pareto fronts and their distribution.

Progressive diversity management in evolutionary multiobjective optimisation

Convergence towards, and diversity across the Pareto-optimal front are the two main requirements when optimising a multiobjective optimisation problem (MOP) with conflicting objectives. Most established multiobjective evolutionary algorithms (MOEAs) have mechanisms that address these requirements. However, in many-objective optimisation, where the number of objectives is greater than 2 or 3, it has been found that these two requirements can conflict with one another, introducing problems such as dominance resistance and speciation. In this study, a previously introduced diversity management mechanism is deployed within a Progressive Preference Articulation (PPA) technique to optimise an 8-objective real-world problem of aircraft control system design. This paper illustrates the effective application of the new diversity management mechanism used in conjunction with the PPA technique when optimising a multiobjective real-world engineering problem.

A Convergence Acceleration Technique for Multiobjective Optimisation

A memetic algorithm which addresses the requirement for solutions convergence towards the Pareto front of a multiobjective optimisation problem is discussed. The memetic algorithm is designed by incorporating a Convergence Accelerator Operator (CAO) in existing algorithms for evolutionary multiobjective optimisation. The discussed convergence accelerator works by suggesting improved solutions in objective space and using neural network mapping schemes to predict the corresponding solution points in decision variable space. Two leading multiobjective evolutionary algorithms have been hybridised through introduction of the CAO and tested on a variety of recognised test problems. These test problems consisted of convex, concave and discontinuous test functions, with numbers of objectives ranging from two to eight. In all cases introduction of the CAO led to improved convergence for comparable numbers of function evaluations.