Search

Diagnosis and prognosis of cancer are informed by the architecture inherent in cancer patient tissue sections. This architecture is typically identified by pathologists, yet advances in computational image analysis facilitate quantitative assessment of this structure. In this article, we develop a spatial point process approach to describe patterns in cell distribution within tissue samples taken from colorectal cancer (CRC) patients. In particular, our approach is centered on the Palm intensity function. This leads to taking an approximate-likelihood technique in fitting point processes models. We consider two Neyman-Scott point processes and a void process, fitting these point process models to the CRC patient data. We find that the parameter estimates of these models may be used to quantify the spatial arrangement of cells. Importantly, we observe characteristic differences in the spatial arrangement of cells between patients who died from CRC and those alive at follow up.

// Hilal S. Khalil 1 , Simon P. Langdon 3 , Alexey Goltsov 1 , Tero Soininen 1 , David J. Harrison 4 , James Bown 2 , Yusuf Y. Deeni 1 1 Division of Science, School of Science, Engineering and Technology, Abertay University, Dundee, DD1 1HG, United Kingdom 2 Division of Computing and Mathematics, School of Arts, Media, and Computer Games, Abertay University, Dundee, DD1 1HG, United Kingdom 3 Division of Pathology, Institute of Genetics and Molecular Medicine, University of Edinburgh, Western General Hospital, Edinburgh, EH4 2XU, United Kingdom 4 School of Medicine, University of St Andrews, St Andrews, KY16 9TF, United Kingdom Correspondence to: Yusuf Y. Deeni, email: y.deeni@abertay.ac.uk Keywords: Pertuzumab/Trastuzumab, NRF2, ROS, HER2-targeted, cancer-immunotherapeutics Received: June 29, 2016 Accepted: September 21, 2016 Published: October 04, 2016 ABSTRACT Nuclear erythroid related factor-2 (NRF2) is known to promote cancer therapeutic detoxification and crosstalk with growth promoting pathways. HER2 receptor tyrosine kinase is frequently overexpressed in cancers leading to uncontrolled receptor activation and signaling. A combination of HER2 targeting monoclonal antibodies shows greater anticancer efficacy than the single targeting antibodies, however, its mechanism of action is largely unclear. Here we report novel actions of anti-HER2 drugs, Trastuzumab and Pertuzumab, involving NRF2. HER2 targeting by antibodies inhibited growth in association with persistent generation of reactive oxygen species (ROS), glutathione (GSH) depletion, reduction in NRF2 levels and inhibition of NRF2 function in ovarian cancer cell lines. The combination of antibodies produced more potent effects than single antibody alone; downregulated NRF2 substrates by repressing the Antioxidant Response (AR) pathway with concomitant transcriptional inhibition of NRF2 . We showed the antibody combination produced increased methylation at the NRF2 promoter consistent with repression of NRF2 antioxidant function, as HDAC and methylation inhibitors reversed such produced transcriptional effects. These findings demonstrate a novel mechanism and role for NRF2 in mediating the response of cancer cells to the combination of Trastuzumab and Pertuzumab and reinforce the importance of NRF2 in drug resistance and as a key anticancer target.

Diagnosis and prognosis of cancer are informed by the architecture inherent in cancer patient tissue sections. This architecture is typically identified by pathologists, yet advances in computational image analysis facilitate quantitative assessment of this structure. In this article, we develop a spatial point process approach to describe patterns in cell distribution within tissue samples taken from colorectal cancer (CRC) patients. In particular, our approach is centered on the Palm intensity function. This leads to taking an approximate-likelihood technique in fitting point processes models. We consider two Neyman-Scott point processes and a void process, fitting these point process models to the CRC patient data. We find that the parameter estimates of these models may be used to quantify the spatial arrangement of cells. Importantly, we observe characteristic differences in the spatial arrangement of cells between patients who died from CRC and those alive at follow up.

Cells are constantly exposed to Reactive Oxygen Species (ROS) produced both endogenously to meet physiological requirements and from exogenous sources. While endogenous ROS are considered as important signalling molecules, high uncontrollable ROS are detrimental. It is unclear how cells can achieve a balance between maintaining physiological redox homeostasis and robustly activate the antioxidant system to remove exogenous ROS. We have utilised a Systems Biology approach to understand how this robust adaptive system fulfils homeostatic requirements of maintaining steady-state ROS and growth rate, while undergoing rapid readjustment under challenged conditions. Using a panel of human ovarian and normal cell lines, we experimentally quantified and established interrelationships between key elements of ROS homeostasis. The basal levels of NRF2 and KEAP1 were cell line specific and maintained in tight correlation with their growth rates and ROS. Furthermore, perturbation of this balance triggered cell specific kinetics of NRF2 nuclear-cytoplasmic relocalisation and sequestration of exogenous ROS. Our experimental data were employed to parameterise a mathematical model of the NRF2 pathway that elucidated key response mechanisms of redox regulation and showed that the dynamics of NRF2-H2O2 regulation defines a relationship between half-life, total and nuclear NRF2 level and endogenous H2O2 that is cell line specific.

We explore the role of narratives of complex systems in anti-cancer drug design. We set out the value of narratives relating to cancer in promoting awareness of risky behaviour and in supporting decision-making regarding treatment options. We present cancer as a dysregulated, complex system that has emergent behaviours at multiple scales, and is governed by dynamical spatio-temporal processes. We show that this system changes structure and function in response to anti-cancer drugs, and explain that these changes are sufficiently complex to impede effective drug design. We pose what narrative might offer to support the process of drug design, providing an example of work done to date that might serve as a foundation for narrating complexity. We suggest ways of using this work combined with that of others to begin to consider narrating drug design.

NF-E2-related factor 2 (NRF2) regulates the transcription of a battery of metabolic and cytoprotective genes. NRF2 and epidermal growth factor receptors (EGFRs/HERs) are regulators of cellular proliferation and determinants of cancer initiation and progression. NRF2 and HERs confer cancers with resistance to several therapeutic agents. Nevertheless, there is limited understanding of the regulation of HER expression and activation and the link between NRF2 and HER signalling pathways. We show that NRF2 regulates both basal and inducible expression ofHER1, as treatment of ovarian cancer cells (PEO1, OVCAR3, and SKOV3) with NRF2 activator tBHQ inducingHER1, while inhibition of NRF2 by siRNA knockdown or with retinoid repressesHER1. Furthermore, treatment of cells with tBHQ increased total and phosphorylated NRF2, HER1, and AKT levels and compromised the cytotoxic effect of lapatinib or erlotinib. Treatment with siRNA or retinoid antagonised the effect of tBHQ on NRF2 and HER1 levels and enhanced the sensitivity of ovarian cancer cells to lapatinib or erlotinib. Pharmacological or genetic inhibition of NRF2 and/or treatment with lapatinib or erlotinib elevated cellular ROS and depleted glutathione. This extends the understanding of NRF2 and its regulation of HER family receptors and opens a strategic target for improving cancer therapy.

We present CoSMoS as a way of structuring thinking on how to reuse parts of an existing model and simulation in a new model and its implementation. CoSMoS provides a lens through which to consider, post-implementation, the assumptions made during the design and implementation of a software simulation of physical interactions in the formation of vascular structures from endothelial cells. We show how the abstract physical model and its software implementation can be adapted for a different problem: the growth of cancer cells under varying environmental perturbations. We identify the changes that must be made to adapt the model to its new context, along with the gaps in our knowledge of the domain that must be filled by wet-lab experimentation when recalibrating the model. Through parameter exploration, we identify the parameters that are critical to the dynamic physical structure of the modelled tissue, and we calibrate these parameters using a series of in vitro experiments. Drawing inspiration from the CoSMoS project structure, we maintain confidence in the repurposed model, and achieve a satisfactory degree of model reuse within our in silico experimental system.

This work was supported by grants from the Northwood Trust (AG, JB), personal support to AG from Scottish Informatics and Computer Science Alliance (SICSA), and EU support for the Concerted Action CASyM (DJH). The phosphatidylinositide 3-kinases (PI3K) and mammalian target of rapamycin-1 (mTOR1) are two key targets for anti-cancer therapy. Predicting the response of the PI3K/AKT/mTOR1 signalling pathway to targeted therapy is made difficult because of network complexities. Systems biology models can help explore those complexities but the value of such models is dependent on accurate parameterisation. Motivated by a need to increase accuracy in kinetic parameter estimation, and therefore the predictive power of the model, we present a framework to integrate kinetic data from enzyme assays into a unified enzyme kinetic model. We present exemplar kinetic models of PI3K and mTOR1, calibrated on in vitro enzyme data and founded on Michaelis-Menten (MM) approximation. We describe the effects of an allosteric mTOR1 inhibitor (Rapamycin) and ATP-competitive inhibitors (BEZ2235 and LY294002) that show dual inhibition of mTOR1 and PI3K. We also model the kinetics of phosphatase and tensin homolog (PTEN), which modulates sensitivity of the PI3K/AKT/mTOR1 pathway to these drugs. Model validation with independent data sets allows investigation of enzyme function and drug dose dependencies in a wide range of experimental conditions. Modelling of the mTOR1 kinetics showed that Rapamycin has an IC50 independent of ATP concentration and that it is a selective inhibitor of mTOR1 substrates S6K1 and 4EBP1: it retains 40% of mTOR1 activity relative to 4EBP1 phosphorylation and inhibits completely S6K1 activity. For the dual ATP-competitive inhibitors of mTOR1 and PI3K, LY294002 and BEZ235, we derived the dependence of the IC50 on ATP concentration that allows prediction of the IC50 at different ATP concentrations in enzyme and cellular assays. Comparison of the drug effectiveness in enzyme and cellular assays showed that some features of these drugs arise from signalling modulation beyond the on-target action and MM approximation and require a systems-level consideration of the whole PI3K/PTEN/AKT/mTOR1 network in order to understand mechanisms of drug sensitivity and resistance in different cancer cell lines. We suggest that using these models in systems biology investigation of the PI3K/AKT/mTOR1 signalling in cancer cells can bridge the gap between direct drug target action and the therapeutic response to these drugs and their combinations. Postprint

Although the theoretical possibility of oscillations in MAPK signalling has long been described, experimental validation has proven more elusive. In this study we observed oscillations in MAPK and PI3K signalling in breast cancer cells in response to epidermal growth factor receptor-family stimulation. Using systems level analysis with a kinetic model, we demonstrate that receptor amplification, loss of transcriptional feedback, or pathway crosstalk, are responsible for oscillations in MAPK and PI3K signalling. Transcriptional profiling reveals architectural motifs likely to be responsible for feedback control of oscillations. Overexpression of the HER2 oncogene and inhibition of transcriptional feedback increase the amplitude of oscillations and provide experimental validation of the computational findings.

// James L. Bown 1, 2 , Mark Shovman 2, 4 , Paul Robertson 2 , Andrei Boiko 1, 2 , Alexey Goltsov 1 , Peter Mullen 3 , David J. Harrison 3 1 School of Science, Engineering and Technology, Abertay University, Dundee, DD1 1HG, UK 2 School of Arts, Media and Computer Games, Abertay University, Dundee, DD1 1HG, UK 3 School of Medicine, University of St Andrews, St Andrews, KY16 9TF, UK 4 Yahoo Labs, Haifa, 31905, Israel Correspondence to: James L. Bown, email: j.bown@abertay.ac.uk David J. Harrison, email: david.harrison@st-andrews.ac.uk Keywords: interactive visualization, systems biology, signaling networks, combination therapy, biomarker discovery Received: January 25, 2016 Accepted: March 31, 2016 Published: May 18, 2016 ABSTRACT Targeted cancer therapy aims to disrupt aberrant cellular signalling pathways. Biomarkers are surrogates of pathway state, but there is limited success in translating candidate biomarkers to clinical practice due to the intrinsic complexity of pathway networks. Systems biology approaches afford better understanding of complex, dynamical interactions in signalling pathways targeted by anticancer drugs. However, adoption of dynamical modelling by clinicians and biologists is impeded by model inaccessibility. Drawing on computer games technology, we present a novel visualization toolkit, SiViT, that converts systems biology models of cancer cell signalling into interactive simulations that can be used without specialist computational expertise. SiViT allows clinicians and biologists to directly introduce for example loss of function mutations and specific inhibitors. SiViT animates the effects of these introductions on pathway dynamics, suggesting further experiments and assessing candidate biomarker effectiveness. In a systems biology model of Her2 signalling we experimentally validated predictions using SiViT, revealing the dynamics of biomarkers of drug resistance and highlighting the role of pathway crosstalk. No model is ever complete: the iteration of real data and simulation facilitates continued evolution of more accurate, useful models. SiViT will make accessible libraries of models to support preclinical research, combinatorial strategy design and biomarker discovery.

Computer simulation can be used to inform in vivo and in vitro experimentation, enabling rapid, low-cost hypothesis generation and directing experimental design in order to test those hypotheses. In this way, in silico models become a scientific instrument for investigation, and so should be developed to high standards, be carefully calibrated and their findings presented in such that they may be reproduced. Here, we outline a framework that supports developing simulations as scientific instruments, and we select cancer systems biology as an exemplar domain, with a particular focus on cellular signalling models. We consider the challenges of lack of data, incomplete knowledge and modelling in the context of a rapidly changing knowledge base. Our framework comprises a process to clearly separate scientific and engineering concerns in model and simulation development, and an argumentation approach to documenting models for rigorous way of recording assumptions and knowledge gaps. We propose interactive, dynamic visualisation tools to enable the biological community to interact with cellular signalling models directly for experimental design. There is a mismatch in scale between these cellular models and tissue structures that are affected by tumours, and bridging this gap requires substantial computational resource. We present concurrent programming as a technology to link scales without losing important details through model simplification. We discuss the value of combining this technology, interactive visualisation, argumentation and model separation to support development of multi-scale models that represent biologically plausible cells arranged in biologically plausible structures that model cell behaviour, interactions and response to therapeutic interventions.

Systems biology approaches that combine experimental data and theoretical modelling to understand cellular signalling network dynamics offer a useful platform to investigate the mechanisms of resistance to drug interventions and to identify combination drug treatments. Extending our work on modelling the PI3K/PTEN/AKT signalling network (SN), we analyse the sensitivity of the SN output signal, phospho-AKT, to inhibition of HER2 receptor. We model typical aberrations in this SN identified in cancer development and drug resistance: loss of PTEN activity, PI3K and AKT mutations, HER2 overexpression, and overproduction of GSK3β and CK2 kinases controlling PTEN phosphorylation. We show that HER2 inhibition by the monoclonal antibody pertuzumab increases SN sensitivity, both to external signals and to changes in kinetic parameters of the proteins and their expression levels induced by mutations in the SN. This increase in sensitivity arises from the transition of SN functioning from saturation to non-saturation mode in response to HER2 inhibition. PTEN loss or PIK3CA mutation causes resistance to anti-HER2 inhibitor and leads to the restoration of saturation mode in SN functioning with a consequent decrease in SN sensitivity. We suggest that a drug-induced increase in SN sensitivity to internal perturbations, and specifically mutations, causes SN fragility. In particular, the SN is vulnerable to mutations that compensate for drug action and this may result in a sensitivity-to-resistance transition. The combination of HER2 and PI3K inhibition does not sensitise the SN to internal perturbations (mutations) in the PI3K/PTEN/AKT pathway: this combination treatment provides both synergetic inhibition and may prevent the SN from acquired mutations causing drug resistance. Through combination inhibition treatments, we studied the impact of upstream and downstream interventions to suppress resistance to the HER2 inhibitor in the SN with PTEN loss. Comparison of experimental results of PI3K inhibition in the PTEN upstream pathway with PDK1 inhibition in the PTEN downstream pathway shows that upstream inhibition abrogates resistance to pertuzumab more effectively than downstream inhibition. This difference in inhibition effect arises from the compensatory mechanism of an activation loop induced in the downstream pathway by PTEN loss. We highlight that drug target identification for combination anti-cancer therapy needs to account for the mutation effects on the upstream and downstream pathways.

The fungi comprise a separate kingdom of life and epitomise the indeterminate growth form. Very little is known about the factors that influence the nature of fungal diversity and the link between individual behaviour and the structure and function of fungal communities is particularly poorly understood. Here, we present a theoretical framework that is capable of elucidating this link. An individual-based model for fungal community dynamics is introduced that has been developed from a physiologically based model for the fungal phenotype. The model is used to explore the role of individual interactions, the production of an external inhibitor field and the quality of the external environment on the structure and diversity of the resulting community. We show that traits relating to growth rate, autophagic behaviour and the production of inhibitors are key in influencing the success of a particular genotype in a community. The species richness increases with the amount of available resource. This is the first model of fungal community dynamics that introduces the concept of a biomass-based abundance distribution function that can be described by the log-normal form which typically corresponds to communities in equilibrium. The species abundance curve was stable to changes in the relative location of inocula, although the ranked abundance of the individuals was not. We present the first attempt to identify the traits that affect the form of that curve. Future studies should examine the role of environmental heterogeneity and spore dispersal.

Overcoming de novo and acquired resistance to anticancer drugs that target signaling networks is a formidable challenge for drug design and effective cancer therapy. Understanding the mechanisms by which this resistance arises may offer a route to addressing the insensitivity of signaling networks to drug intervention and restore the efficacy of anticancer therapy. Extending our recent work identifying PTEN as a key regulator of Herceptin sensitivity, we present an integrated theoretical and experimental approach to study the compensatory mechanisms within the PI3K/PTEN/AKT signaling network that afford resistance to receptor tyrosine kinase (RTK) inhibition by anti-HER2 monoclonal antibodies. In a computational model representing the dynamics of the signaling network, we define a single control parameter that encapsulates the balance of activities of the enzymes involved in the PI3K/PTEN/AKT cycle. By varying this control parameter we are able to demonstrate both distinct dynamic regimes of behavior of the signaling network and the transitions between those regimes. We demonstrate resistance, sensitivity, and suppression of RTK signals by the signaling network. Through model analysis we link the sensitivity-to-resistance transition to specific compensatory mechanisms within the signaling network. We study this transition in detail theoretically by variation of activities of PTEN, PI3K, AKT enzymes, and use the results to inform experiments that perturb the signaling network using combinatorial inhibition of RTK, PTEN, and PI3K enzymes in human ovarian carcinoma cell lines. We find good alignment between theoretical predictions and experimental results. We discuss the application of the results to the challenges of hypersensitivity of the signaling network to RTK signals, suppression of drug resistance, and efficacy of drug combinations in anticancer therapy.

This contribution, based on a Special Interest Group session held during IMC9, focuses on physiological based models of filamentous fungal colony growth and interactions. Fungi are known to be an important component of ecosystems, in terms of colony dynamics and interactions within and between trophic levels. We outline some of the essential components necessary to develop a fungal ecology: a mechanistic model of fungal colony growth and interactions, where observed behaviour can be linked to underlying function; a model of how fungi can cooperate at larger scales; and novel techniques for both exploring quantitatively the scales at which fungi operate; and addressing the computational challenges arising from this highly detailed quantification. We also propose a novel application area for fungi which may provide alternate routes for supporting scientific study of colony behaviour. This synthesis offers new potential to explore fungal community dynamics and the impact on ecosystem functioning.

NF-E2 related factor-2 (NRF2) is an essential transcription factor for multiple genes encoding antioxidants and detoxification enzymes. NRF2 is implicated in promoting cancer therapeutic resistance by its detoxification function and crosstalk with proproliferative pathways. However, the exact mechanism of this intricate connectivity between NRF2 and growth factor induced proliferative pathway remains elusive. Here, we have demonstrated that pharmacological activation of NRF2 bytert-butylhydroquinone (tBHQ) upregulates the HER family receptors, HER2 and HER3 expression, elevates pAKT levels, and enhances the proliferation of ovarian cancer cells. Preactivation of NRF2 also attenuates the combined growth inhibitory effects of HER2 targeting monoclonal antibodies, Pertuzumab and Trastuzumab. Further, tBHQ caused transcriptional induction of HER2 and HER3, while SiRNA-mediated knockdown of NRF2 prevented this and further caused transcriptional repression and enhanced cytotoxicity of the HER2 inhibitors. Hence, NRF2 regulates both HER2 and HER3 receptors to influence cellular responses to HER2 targeting monoclonal antibodies. This deciphered crosstalk mechanism reinforces the role of NRF2 in drug resistance and as a relevant anticancer target.

A central aim in ecology is to understand the relation between organism diversity and ecosystem functioning. We investigate this relation using a generic individual-based modelling framework described in part 1, in which individuals within a community are characterised by physiological traits and interact within a spatially structured environment. We explore the effect of intraspecific variation among individuals on community-scale productivity in a range of homogeneous and heterogeneous environments. We show that diversity among individuals has a significant, and in most but not all cases positive, impact on community productivity. At low levels of resource the persisting plants giving highest productivity have slowest uptake rates and longest times to reproduction. In contrast, at high levels of resource the persisting plants giving highest productivity have highest uptake levels and shortest time to reproduction. Thus, the individuals’ trait distributions defining community functioning emerge, after many iterations of the model, from the surviving individuals. We, therefore, show that different mechanisms, as evidenced in the surviving individuals’ traits, increase community productivity in different environmental contexts, and so demonstrate the importance of accounting for diversity at the scale of the individual. Consequently, progress may be made in linking diversity and function by considering small, manageable systems and making physiological measurements on individuals.

Biodiversity is generally accepted to include both within species and between species vari- ation. Consequently, the contribution to the functioning of ecosystems of variation among individuals should be accounted for. However, very little is known about patterns of diver- sity below the species level, and less still about the patterns of individual diversity and their relation to ecosystem context and community function. We present a model for the dynamics of individuals that is physiologically based and spatially explicit. Individuals are defined in terms of measurable parameters that relate environmental context to phenotype and in this sense define the genotype. Estimates for the variation in the parameter values are obtained from experiments conducted on the species Rumex acetosa. Simulations are performed to predict the form of the relative abundance distribution, and the dependence of the predicted number of coexisting genotypes on patch area (the genotype-area curve). We find that the predicted forms of the abundance distribution and genotype-area curve are indistinguishable from those measured at the species level. Furthermore, we identify the importance of physiological trade-offs at the individual level in promoting diversity and the sensitivity of genotype richness to the degree of similarity of individuals in the community. The extension of the model to include inter-species variation is discussed.

Using an existing fungal growth model that captures the physiological processes of vegetative growth and development of a fungal colony, and in particular incorporates, for the first time, a recycling of biomass mechanism, we explore the effects of recycling in various environmental contexts. Here we test whether resource density thresholds exist, below which finite colony expansion occurs, in three dimensions based on the number of randomly removed resource sites. We then test the effect of recycling on resource density thresholds. Modelled soil structure, derived from experiments, is combined with the fungal growth model. The effect of recycling on foraging efficiency is investigated for resource distributed homogeneously and heterogeneously throughout the modelled soil structure. The simulated results show that resource density thresholds do exist in three dimensions and that the recycling mechanism decreases the threshold value. Our results indicate that recycling promotes persistence and a recycling mechanism is crucial for those fungi that reside in a resource patchy and limited environment.

Fungi are one of the most important and widespread components of the biosphere, and are essential for the growth of over 90% of all vascular plants. Although they are a separate kingdom of life, we know relatively little about the origins of their ubiquitous existence. This reflects a wider ignorance arising from their status as indeterminate organisms epitomized by extreme phenotypic plasticity that is essential for survival in complex environments. Here we show that the fungal phenotype may have its origins in the defining characteristic of indeterminate organisms, namely their ability to recycle locally immobilized internal resources into a mobilized form capable of being directed to new internal sinks. We show that phenotype can be modelled as an emergent phenomenon resulting from the interplay between simple local processes governing uptake and remobilization of internal resources, and macroscopic processes associated with their transport. Observed complex growth forms are reproduced and the sensitive dependence of phenotype on environmental context may be understood in terms of nonlinearities associated with regulation of the recycling apparatus.

The receptor tyrosine kinases (RTKs) are key drivers of cancer progression and targets for drug therapy. A major challenge in anti-RTK treatment is the dependence of drug effectiveness on co-expression of multiple RTKs which defines resistance to single drug therapy. Reprogramming of the RTK network leading to alteration in RTK co-expression in response to drug intervention is a dynamic mechanism of acquired resistance to single drug therapy in many cancers. One route to overcome this resistance is combination therapy. We describe the results of a joint in silico, in vitro, and in vivo investigations on the efficacy of trastuzumab, pertuzumab and their combination to target the HER2 receptors. Computational modelling revealed that these two drugs alone and in combination differentially suppressed RTK network activation depending on RTK co-expression. Analyses of mRNA expression in SKOV3 ovarian tumour xenograft showed up-regulation of HER3 following treatment. Considering this in a computational model revealed that HER3 up-regulation reprograms RTK kinetics from HER2 homodimerisation to HER3/HER2 heterodimerisation. The results showed synergy of the trastuzumab and pertuzumab combination treatment of the HER2 overexpressing tumour can be due to an independence of the combination effect on HER3/HER2 composition when it changes due to drug-induced RTK reprogramming. Publisher PDF

The study of patterns in living diversity is driven by the desire to find the universal rules that underlie the organization of ecosystems. The relative abundance distribution, which characterizes the total number and abundance of species in a community, is arguably the most fundamental measure in ecology. Considerable effort has been expended in striving for a general theory that can explain the form of the distribution. Despite this, a mechanistic understanding of the form in terms of physiological and environmental parameters remains elusive. Recently, it has been proposed that space plays a central role in generating the patterns of diversity. Here we show that an understanding of the observed form of the relative abundance distribution requires a consideration of how individuals pack in time. We present a framework for studying the dynamics of communities which generalizes the prevailing species-based approach to one based on individuals that are characterized by their physiological traits. The observed form of the abundance distribution and its dependence on richness and disturbance are reproduced, and can be understood in terms of the trade-off between time to reproduction and fecundity.

A stochastic cellular automaton for modelling the dynamics of a two-species fungal microcosm is presented. The state of each cell in the automaton depends on the state of a predefined neighbourhood via a set of conditional probabilities derived from experiments conducted on pairwise combinations of species. The model is tested by detailed comparison with larger-scale experimental microcosms. By employing different hypotheses which relate the pairwise data to the conditional probabilities in the model, the nature of the local and non-local interactions in the community is explored. The hypothesis that the large-scale dynamics are a consequence of independent interactions between species in a local neighbourhood can be excluded at the 5% significance level. The form of the interdependencies is determined and it is shown that the outcome of the interactions at the local neighbourhood-scale depends on the community-scale patterning of individuals. The dynamics of the microcosm are therefore an emergent property of the system of interacting mycelia that cannot be deduced from a study of the components in isolation.

The community dynamics of two- and three-fungal species interactions derived for a tessellated agar model system are described. The microcosm allows for the varied prescription of: (1) the number of fungal species interacting; (2) the spatial configuration (patchiness) of the distribution of individuals; (3) the magnitude of scale of spatial occupation by different fungal individuals; and (4) the operation of antagonistic mechanisms based on contact or longer range diffusible components. Stepwise logistic regressions for two-species interactions are used to inform the design of the multi-species interaction tessellations. The model prescribes and investigates complex parameters, such as spatiotemporal heterogeneity and microcosm scale (e.g. population patchiness and crossing times). Data are quantified as proportion, interface class and state transition class of viable fungal species. Spatiotemporal heterogeneity is represented using a novel application of principal component analysis which shows good intuitive agreement with visual assessment of the interaction outcome patterns, and allows effective comparison of the data as a whole. The model demonstrates the influence of the complex and coordinated behaviour of fungal mycelia on community development: interaction outcome of three-species interactions cannot be directly extrapolated from the relevant binary component interactions; interaction outcomes of the multi-species tessellations appears to be neither random nor fully deterministic; a degree of stochasticity is apparent in all tessellation arrangements; the smaller scale tessellations produce more consistent interaction outcome results, probably because experimental scale affects the duration of transient behaviour; and different initial spatial configurations of inoculum (irrespective of inoculum quantity or proportion) influence community development and reproducibility.

This work was supported by grants from Breakthrough Breast Cancer and Scottish Funding Council (SRDG), and personal support to Alexey Goltsov from Scottish Informatics and Computer Science Alliance (SICSA) and to James Bown from The Northwood Trust. Drug resistance, de novo and acquired, pervades cellular signaling networks (SNs) from one signaling motif to another as a result of cancer progression and/or drug intervention. This resistance is one of the key determinants of efficacy in targeted anti-cancer drug therapy. Although poorly understood, drug resistance is already being addressed in combination therapy by selecting drug targets where SN sensitivity increases due to combination components or as a result of de novo or acquired mutations. Additionally, successive drug combinations have shown low resistance potential. To promote a rational, systematic development of combination therapies, it is necessary to establish the underlying mechanisms that drive the advantages of combination therapies, and design methods to determine drug targets for combination regimens. Based on a joint systems analysis of cellular SN response and its sensitivity to drug action and oncogenic mutations, we describe an in silico method to analyze the targets of drug combinations. Our method explores mechanisms of sensitizing the SN through a combination of two drugs targeting vertical signaling pathways. We propose a paradigm of SN response customization by one drug to both maximize the effect of another drug in combination and promote a robust therapeutic response against oncogenic mutations. The method was applied to customize the response of the ErbB/PI3K/PTEN/AKT pathway by combination of drugs targeting HER2 receptors and proteins in the down-stream pathway. The results of a computational experiment showed that the modification of the SN response from hyperbolic to smooth sigmoid response by manipulation of two drugs in combination leads to greater robustness in therapeutic response against oncogenic mutations determining cancer heterogeneity. The application of this method in drug combination co-development suggests a combined evaluation of inhibition effects together with the capability of drug combinations to suppress resistance mechanisms before they become clinically manifest. Publisher PDF

New methods for inferring data-consistent, selfreconfigurable power-law models from time series data are required and developed. These novel methods may be categorised into two broad groups, namely: straightforward (or direct) inference methods based on power-law models; and a jacobian based indirect inference method. The direct method involves applying direct means to infer a power-law model from time series data. The indirect method, however, uses a new system identification method to first infer a jacobian model as instant and temporal solution to the inverse problem before recasting the inferred jacobian model to corresponding power-law model using our newly developed recast technique. The recast method, in addition to normal behaviour, also provides a novel analytical technique for integrating power-law and jacobian models together. The modelling approach we have developed extends previous work on matrix-based network inference to model interoperability and multiple model transformation in terms of finding two distinct models (solutions) to an inverse problem.

SUMMARY In cancer, morphological assessment of histological tissue samples is a fundamental part of both diagnosis and prognosis. Image analysis offers opportunities to support that assessment through quantitative metrics of morphology. Generally, morphometric analysis is carried out on two- dimensional tissue section data and so only represents a small fraction of any tumour. We present a novel application of three-dimensional (3D) morphometrics for 3D imaging data obtained from tumours grown in a culture model. Minkowski functionals, a set of measures that characterise geometry and topology in n-dimensional space, are used to quantify tumour topology in the absence of and in response to therapeutic intervention. These measures are used to stratify the morphological response of tumours to therapeutic intervention. Breast tumours are characterised by estrogen receptor (ER) status, human epidermal growth factor receptor (HER)2 status and tumour grade. Previously, we have shown that ER status is associated with tumour volume in response to tamoxifen treatment ex vivo. Here, HER2 status is found to predict the changes in morphology other than volume as a result of tamoxifen treatment ex vivo. Finally, we show the extent to which Minkowski functionals might be used to predict tumour grade. Minkowski functionals are generalisable to any 3D data set, including in vivo and cellular systems. This quantitative topological analysis can provide a valuable link among biomarkers, drug intervention and tumour morphology that is complementary to existing, non-morphological measures of tumour response to intervention and could ultimately inform patient treatment.

Dynamic processes in biological systems may be profiled by measuring system properties over time. One way of representing such time series data is through weighted interaction networks, where the nodes in the network represent the measurables and the weighted edges represent interactions between any pair of nodes. Construction of these network models from time series data may involve seeking a robust data-consistent and time-invariant model to approximate and describe system dynamics. Many problems in mathematics, systems biology and physics can be recast into this form and may require finding the most consistent solution to a set of first order differential equations. This is especially challenging in cases where the number of data points is less than or equal to the number of measurables. We present a novel computational method for network reconstruction with limited time series data. To test our method, we use artificial time series data generated from known network models. We then attempt to reconstruct the original network from the time series data alone. We find good agreement between the original and predicted networks.

In a world of ever-increasing and newly discovered complexities, and rapidly expanding data sets describing man-made and natural phenomena, information visualization offers a means of structuring and enabling interpretation of these data in the context of that complexity. Advances in graphics hardware, art asset pipelines and parallelized computational platforms offer unprecedented potential. However, harnessing this potential to good effect is challenging and requires the integration of skills from the arts and social sciences to support scientific endeavor in the physical and life sciences. Here, we consider those skills and describe four case studies that highlight interoperation among disciplines at this arts-science-social science interface.

Cancer is a complex and heterogeneous disease, not only at a genetic and biochemical level, but also at a tissue, organism, and population level. Multiple data streams, from reductionist biochemistry in vitro to high-throughput "-omics" from clinical material, have been generated with the hope that they encode useful information about phenotype and, ultimately, tumour behaviour in response to drugs. While these data stand alone in terms of the biology they represent, there is the enticing prospect that if incorporated into systems biology models, they can help understand complex systems behaviour and provide a predictive framework as an additional tool in understanding how tumours change and respond to treatment over time. Since these biological data are heterogeneous and frequently qualitative rather than quantitative, at the present time a single systems biology approach is unlikely to be effective; instead, different computational and mathematical approaches should be tailored to different types of data, and to each other, in order to test and re-test hypotheses. In time, these models might converge and result in usable tractable models which accurately represent human cancer. Likewise, biologists and clinicians need to understand what the requirements of systems biology are so that compatible data are produced for computational modelling. In this review, we describe some theoretical approaches (data-driven and process-driven) and experimental methodologies which are being used in cancer research and the clinical context where they might be applied.

Research based on computer simulations, especially that conducted through agent-based experimentation, is often criticised for not being a reliable source of information - the simulation software can hide errors or flawed designs that inherently bias results. Consequently, the academic community shows both enthusiasm and lack of trust for such approaches. In order to gain confidence is using engineered systems, domains such as Safety Critical Systems employ structured argumentation techniques as means of explicitly relating claims to evidence - in other words, requirements to deliverables. We argue here that structured argumentation should be used in the development and validation process of simulation-driven research. Making use of the Goal Structuring Notation, we provide insights into how more trustworthy outcomes can be obtained through argumentation-driven validation.

In logic, an argument is valid if true premises cannot be associated with false conclusions. However, this criterion is hardly applicable to real-world complex systems, hence it needs to be weakened: critical systems engineering, for example, requires acceptability rather than truth. Validity cannot be assumed, nor is it transitive; it has to be demonstrated within a context and related to a specific set of requirements. We present here work directed towards developing a structured argument of validity for a complex systems simulation - part of a plant ecology case-study. Firstly, we define the difficulties of claiming validity for complex systems simulations. We then proceed to analysing the spectrum of assumptions and abstractions that were involved in constructing the biological model and simulation, and how these influence the scope of our validity claim. Finally, we provide an initial structured argument for the simulation's validity, using critical systems techniques.

Prevention and control of Healthcare Associated Infections (HAIs) has become a high priority for most healthcare organizations. Mathematical models can provide insights into the dynamics of nosocomial infections and help to evaluate the effect of infection control measures. The model presented in this paper adopts an individual-based and stochastic approach to investigate MRSA outbreaks in a hospital ward. A computer simulation was implemented to analyze the dynamics of the system associated with the spread of the infection and to carry out studies on space and personnel management. This study suggests that a strict spatial cohorting might be ineffective, if it is not combined with personnel cohorting.

We describe the concept of a cognitive network and propose that ecosystems of co-existing networks which are globally energy efficient while providing the expected quality of service can be realised by exploiting two mechanisms which occur in biological systems; symbiosis and co-existence.

Cancer is a complex and heterogeneous disease, not only at a genetic and biochemical level, but also at a tissue, organism, and population level. Multiple data streams, from reductionist biochemistry in vitro to high-throughput "-omics" from clinical material, have been generated with the hope that they encode useful information about phenotype and, ultimately, tumour behaviour in response to drugs. While these data stand alone in terms of the biology they represent, there is the enticing prospect that if incorporated into systems biology models, they can help understand complex systems behaviour and provide a predictive framework as an additional tool in understanding how tumours change and respond to treatment over time. Since these biological data are heterogeneous and frequently qualitative rather than quantitative, at the present time a single systems biology approach is unlikely to be effective; instead, different computational and mathematical approaches should be tailored to different types of data, and to each other, in order to test and re-test hypotheses. In time, these models might converge and result in usable tractable models which accurately represent human cancer. Likewise, biologists and clinicians need to understand what the requirements of systems biology are so that compatible data are produced for computational modelling. In this review, we describe some theoretical approaches (data-driven and process-driven) and experimental methodologies which are being used in cancer research and the clinical context where they might be applied.

Consequences of initial spatial organisation of model fungal communities upon their spatio-temporal development were investigated. Dynamics of prescribed two- and three-species 'communities' developing on tessellated agar tile model systems were analysed in terms of literal maps, principal component analyses, or as the proportion of species extant within tiles. It was established that for two-species interactions of equal patch size, large-scale (i.e. many constituent tiles) behaviour could be extrapolated from the relevant small-scale (i.e. pairs of tiles) interactions. However, relative patch sizes (scale) of species within tessellations influenced the times taken by individuals to colonise tiles and, hence, temporal behaviour of the system. Outcome of arrangements involving three species of equal patch size and inoculum potential, and prescribed with different mixing patterns, could not be directly extrapolated by reference to the outcome of pair-wise interactions between constituent species. Three-species arrangements attempt to limit assembly of lateral aggregates of individuals (patch size) and hence any effects of tile colonisation times, so as to reveal effects of nearest neighbour context within the complex community. Such arrangements indicate that spatial configuration of inoculum influences community development and reproducibility. They also suggest that spatial distribution of species affects persistence of individuals, which would otherwise be expected to be eliminated from the system. Two-species interactions appeared generally more reproducible than those comprising three species, and the sensitivity of fungal community development to temperature was not solely associated with influence on colony extension rate.

The core problem of the evaluation of Information Visualisation is that the end product of visualisation – the comprehension of the information from the data – is difficult to measure objectively. This paper outlines a description of visualisation comprehension based on two existing theories of perception: Principles of Perceptual Organisation and the Reverse Hierarchy Theory. The resulting account of the processes involved in visualisation comprehension enables evaluation that is not only objective, but also non-comparative, providing an absolute efficiency classification. Finally, as a sample application of this approach, an experiment studying the benefits of interactivity in 3D scatterplots is presented.

Ataxia-telangiectasia mutated (ATM) is known to play a central role in effecting the DNA damage response that protects somatic cells from potentially harmful mutations, and in this role it is a key anti-cancer agent. However, it also promotes repair of therapeutic damage (e.g. radiotherapy) and so frustrates the efficacy of some treatments. A better understanding of the mechanisms of ATM regulation is therefore important both in prevention and treatment of disease. While progress has been made in elucidating the key signal transduction pathways that mediate damage response in somatic cells, relatively little is known about whether these function similarly in pluripotent embryonic stem (ES) cells where ATM is also implicated in our understanding of adult stem cell ageing and in improvements in regenerative medicine. There is some evidence that different mechanisms may operate in ES cells and that our understanding of the mechanisms of ATM regulation is therefore incomplete. We investigated the behaviour of the damage response signalling pathway in mouse ES cells. We subjected the cells to the DNA-damaging agent doxorubicin, a drug that induces double-strand breaks, and measured ATM expression levels. We found that basal ATM gene expression was unaffected by doxorubicin treatment. However, following ATM kinase inhibition using a specific ATM inhibitor, we observed a significant increase in ATM and ataxia-telangiectasia and Rad3 related transcription. We demonstrate the use of a dynamical modelling approach to show that these results cannot be explained in terms of known mechanisms. Furthermore, we show that the modelling approach can be used to identify a novel feedback process that may underlie the anomalies in the data. The predictions of the model are consistent both with our in vitro experiments and with in vivo studies of ATM expression in somatic cells in mice, and we hypothesize that this feedback operates in both somatic and ES cells in vivo . The results point to a possible new target for ATM inhibition that overcomes the restorative potential of the proposed feedback.