Content uploaded by V Anne Smith
Author content
All content in this area was uploaded by V Anne Smith on Sep 28, 2015
Content may be subject to copyright.
A preview of the PDF is not available
Cancer Systems Biology
Abstract and Figures
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.
Figures - uploaded by V Anne Smith
Author content
All figure content in this area was uploaded by V Anne Smith
Content may be subject to copyright.
... The occurrence of cancer requires a series of cellular phenotypic changes [25]. Both genetic and environmental factors are involved in tumorigenesis [26]. The epigenetic characteristics of cancers may differ due to different pathogenic factors. ...
... A better understanding of the biological characteristics of tumors may facilitate the development of cancer treatment. The hallmarks of tumor progression include over-activation of growth signals, insensitivity to growth-inhibitory signals, avoidance of apoptosis, unrestricted replication potential, continuous vascular growth, tissue invasion, and immune reediting [26]. Melanoma is a highly immune-related tumor. ...
... The over-activation of growth signals, such as the ErbB, VEGF, and PI3K signaling pathways [25], leads to tumorigenesis, tumor progression, and metastasis [26]. The ErbB receptors include ERBB1 (EGFR/ HER1), ERBB2 (HER2), ERBB3, and ERBB4. ...
Melanoma is a type of cancer with a relatively poor prognosis. The development of immunotherapy for the treatment of patients with melanoma has drawn considerable attention in recent years. It is of great clinical significance to identify novel promising prognostic biomarkers and to explore their roles in the immune microenvironment. The solute carrier (SLC) superfamily is a group of transporters predominantly expressed on the cell membrane and are involved in substance transport. SLC16A1 is a member of the SLC family, participating in the transport of lactate, pyruvate, amino acids, ketone bodies, etc. The role of SLC16A1 in tumor immunity has been recently elucidated, while its role in melanoma remains unclear. In this study, bioinformatics analysis was performed to explore the role of SLC16A1 in melanoma. The results showed that high SLC16A1 expression was correlated with decreased overall survival in patients with melanoma. The genes co-expressed with SLC16A1 were significantly enriched in metabolic regulation, protein ubiquitination, and substance localization. Moreover, SLC16A1 was correlated with the infiltration of immune cells. In conclusion, SLC16A1 is a robust prognostic biomarker for melanoma and may be used as a novel target in immunotherapy.
... Kinetic modeling using both experimental and mathematical data can now be used to assess tumor biology over time (31). Some neural networks have been in place clinically for several years for the conversion of MRI data into a three-dimensional tumor landscape in order to determine a target area for radiotherapy (40,41). ...
... The ultimate aim of these neural networks is to provide a methodology that can be used to convert biomarker data (and associated aberrant pathway signaling) into a treatment regime, based on a predicted outcome (31). If this were the case, the true nature of tumor biology may be identified, allowing a reduction in the use of inferred cancer dynamics from biomarker analysis (31). ...
... The ultimate aim of these neural networks is to provide a methodology that can be used to convert biomarker data (and associated aberrant pathway signaling) into a treatment regime, based on a predicted outcome (31). If this were the case, the true nature of tumor biology may be identified, allowing a reduction in the use of inferred cancer dynamics from biomarker analysis (31). However, the capacity of any such mathematical model means it is unlikely to be able to describe all parts of the network over space and time due to the amount of biological variation present (31). ...
Glioblastoma (GB) is the most common primary malignant brain tumor, and despite the availability of chemotherapy and radiotherapy to combat the disease, overall survival remains low with a high incidence of tumor recurrence. Technological advances are continually improving our understanding of the disease, and in particular, our knowledge of clonal evolution, intratumor heterogeneity, and possible reservoirs of residual disease. These may inform how we approach clinical treatment and recurrence in GB. Mathematical modeling (including neural networks) and strategies such as multiple sampling during tumor resection and genetic analysis of circulating cancer cells, may be of great future benefit to help predict the nature of residual disease and resistance to standard and molecular therapies in GB.
... To model cancer progression, quantitative relationships and dependencies between functional elements within a cell need to be captured. While these dependencies are traditionally characterized using systems biology models based on ordinary or partial differential equations [26], these quantitative relations are not immediately accessible in many machine learning models. Furthermore, while there has been significant progress in interpretability of machine learning methods [27], they are not yet regularly applied to high-dimensional biological data. ...
Combining multiple types of genomic, transcriptional, proteomic, and epigenetic datasets has the potential to reveal biological mechanisms across multiple scales, and may lead to more accurate models for clinical decision support. Developing efficient models that can derive clinical outcomes from high-dimensional data remains problematical; challenges include the integration of multiple types of omics data, inclusion of biological background knowledge, and developing machine learning models that are able to deal with this high dimensionality while having only few samples from which to derive a model. We developed DeepMOCCA, a framework for multi-omics cancer analysis. We combine different types of omics data using biological relations between genes, transcripts, and proteins, combine the multi-omics data with background knowledge in the form of protein–protein interaction networks, and use graph convolution neural networks to exploit this combination of multi-omics data and background knowledge. DeepMOCCA predicts survival time for individual patient samples for 33 cancer types and outperforms most existing survival prediction methods. Moreover, DeepMOCCA includes a graph attention mechanism which prioritizes driver genes and prognostic markers in a patient-specific manner; the attention mechanism can be used to identify drivers and prognostic markers within cohorts and individual patients.
Author summary
Linking the features of tumors to a prognosis for the patient is a critical part of managing cancer. Many methods have been applied to this problem but we still lack accurate prognostic markers for many cancers. We now have more information than ever before on the state of the cancer genome, the epigenetic changes in tumors, and gene expression at both RNA and protein levels. Here, we address the question of how this data can be used to predict cancer survival and discover which tumor genes make the greatest contribution to the prognosis in individual tumor samples. We have developed a computational model, DeepMOCCA, that uses artificial neural networks underpinned by a large graph constructed from background knowledge concerning the functional interactions between genes and their products. We show that with our method, DeepMOCCA can predict cancer survival time based entirely on features of the tumor at a cellular and molecular level. The method confirms many existing genes that affect survival but for some cancers suggests new genes, either not implicated in survival before or not known to be important in that particular cancer. The ability to predict the important features in individual tumors provided by our method raises the possibility of personalized therapy based on the gene or network dominating the prognosis for that patient.
... Consequently, cancer systems immunology represents a new avenue of interrogation for understanding how the immune system interacts with tumors during tumorigenesis, progression, and treatment. Cancer systems biology and systems immunology have been reviewed elsewhere Faratian, 2010;Suhail et al., 2019;Germain et al., 2011;Vera, 2015;Werner et al., 2014;Korsunsky et al., 2014;Kreeger and Lauffenburger, 2010;Chuang et al., 2010). In this review, we will discuss approaches to the nascent field of cancer systems immunology as well as their potential applications and current limitations. ...
Tumor immunology is undergoing a renaissance due to the recent profound clinical successes of tumor immunotherapy. These advances have coincided with an exponential growth in the development of –omics technologies. Armed with these technologies and their associated computational and modeling toolsets, systems biologists have turned their attention to tumor immunology in an effort to understand the precise nature and consequences of interactions between tumors and the immune system. Such interactions are inherently multivariate, spanning multiple time and size scales, cell types, and organ systems, rendering systems biology approaches particularly amenable to their interrogation. While in its infancy, the field of ‘Cancer Systems Immunology’ has already influenced our understanding of tumor immunology and immunotherapy. As the field matures, studies will move beyond descriptive characterizations toward functional investigations of the emergent behavior that govern tumor-immune responses. Thus, Cancer Systems Immunology holds incredible promise to advance our ability to fight this disease.
... The challenge lies in that cancer is not a unique disease but many different manifestations, even within the tumor itself. For systems biology, each tumor constitutes a system characterized by its cellular heterogeneity, the microenvironment (surrounding tissue, immune system) in which it grows, and its ability to adapt and modify it 1 . That is, a tumor is a unique and variable entity, product of the multiple mutations and epigenetic alterations that occur in some of the thousands of cells in the early stages of malignant transformation. ...
The battle against cancer has advanced tremendously in the last thirty years and the survival rate has doubled. However, it is still difficult to achieve a generalized cure. The challenge is that cancer is not a unique disease; it is about dozens of different manifestations, even within the same tumor location. For systems biology, each solid tumor is a unique system characterized by its cellular heterogeneity, its interaction with the microenvironment in which it grows and develops, and its ability to adapt and modify it. Recent advances in understanding the molecular mechanisms that underlie cancer are transforming the diagnosis and treatment of the disease. In this sense, a growing set of treatments capable of attacking a specific tumor with higher efficiency has been developed, defining a new paradigm: the precision medicine in oncology. Genomics and bioinformatics are two fundamental pillars in this applied field. These technologies generate massive data (Big Data) that require analytical tools and trained personnel for the analysis, integration and transfer of the information to physicians. This presentation describes the concepts of personalized medicine, Big Data, the main advances in genomics and bioinformatics as well as their future perspectives and challenges.
Gastric cancer is among the cancers with the highest mortality in the world, and treatment modalities involve surgery, chemotherapy and radiotherapy. In chemotherapy, the drugs alone or in combination face a dilemma common in this disease: the chemoresistance. Thus, it becomes essential in the clinic, the need to find new protein target or compounds that are capable of maintaining a response for long periods of treatment. Thus, this studied used tools of systems pharmacology, evaluating the interactions between proteins and small compounds from existing proteomic data for the human species. In this context, the study drugs were 5-fluorouracil, capecitabine, oxaliplatin, irinotecan and docetaxel. A network of protein interactions and protein-chemotherapy (PPPI-PCPI, respectively) was obtained, followed by the definition of five sub-networks, representing different biological processes observed. Moreover, centrality analysis were used to search the main network components and their interactions, called bottlenecks , which regulate the flow of information within the network. Thus, the study showed that the drugs currently in clinical use converge to similar biological processes, which would explain the development of chemoresistance in the medium and long term. Furthermore, we identified seven new bottlenecks that represent new target proteins and small compounds able to interfere in the control and communication between other nodes in the network itself. These data could be used to develop new combinations of drugs with the aim of improving the protocols used in chemotherapy of gastric cancer.
Nanoparticles promise to improve the treatment of cancer through their increasingly sophisticated functionalisations and ability to accumulate in certain tumours. Yet recent work has shown that many nanomedicines fail during clinical trial. One issue is the lack of understanding of how nanoparticle designs impact their ability to overcome transport barriers in the body, including their circulation in the blood stream, extravasation into tumours, transport through tumour tissue, internalisation in the targeted cells, and release of their active cargo. Increased computational power, as well as improved multi-scale simulations of tumours, nanoparticles, and the biological transport barriers that affect them, now allow us to investigate the influence of a range of designs in biologically relevant scenarios. This presents a new opportunity for high-throughput, systematic, and integrated design pipelines powered by data and machine learning. With this paper, we review latest results in multi-scale simulations of nanoparticle transport barriers, as well as available software packages, with the aim of focussing the wider research community in building a common computational framework that can overcome some of the current obstacles facing efficient nanoparticle design.
A combination of decades of cancer research and trillions of dollars has helped to exponentially increase our understanding of the molecular processes involved in cancer pathogenesis [1, 2] and to develop new cancer drugs. However, despite progress in diagnosing and treating cancer, these diseases remain one of the leading causes of morbidity and mortality worldwide, responsible for millions of deaths each year [3]. Moreover, we are still faced, on average, with low drug response rates, very serious treatment side effects and questionable survival benefits. In Europe alone, cancer kills around 4000 people every day [4] and costs billions per year [5]. Worldwide, the number of new cancer cases is increasing every year [4], at least partly due to rapidly ageing populations [6], with the number of new cancer cases projected to reach 25 million per year by 2030, and cure rates for many common forms of cancer stagnating [4]. Due to the high number of nonresponders to existing drugs and spiralling costs of new cancer drugs—costs have almost doubled in the last decade, resulting in a dramatic decrease in the number of new drugs—an individualised approach to the diagnosis and treatment of cancer patients is desperately required.
The optimal targeting of cancer requires not only the selection of the target but also the identification of the patients whose cancer depends on the targeted pathway. The objective of this chapter is to give an overview of molecular targets in cancer therapeutics. Targets have been categorized as either established or novel types. Established targets include those against which most currently licensed anticancer drugs were developed and include DNA, microtubules, and nuclear hormone receptors. Novel targets are those under current preclinical and clinical investigation. The section on novel targets emphasizes the relationships of the novel targets to the biological traits of cancer.
Breast cancer is the second most common malignancy for women. Currently, only several prognostic and predictive factors are used clinically for managing breast cancer patients. Recently, systems biology approaches based on high-throughput technologies such as DNA microarrays, mass spectrometry-based proteomics and metabolomics have began to be used to investigate the expression of a wide range of genes and proteins in the dissected breast tumors. Moreover, these expression signatures have been found to provide potential and independent prognostic information in patients diagnosed with breast cancer. Furthermore, these molecular signatures could not only help to identify new therapeutic targets, but also allow physicians to design more effective and targeted therapeutic strategies for achieving better treatment outcomes of breast cancer patients. © 2013 Springer Science+Business Media New York. All rights are reserved.
A new method for sensitivity analysis (SA) of model output is introduced. It is based on the Fourier amplitude sensitivity test (FAST) and allows the computation of the total contribution of each input factor to the output's variance. The term “total” here means that the factor's main effect, as well as all the interaction terms involving that factor, are included. Although computationally different, the very same measure of sensitivity is offered by the indices of Sobol'. The main advantages of the extended FAST are its robustness, especially at low sample size, and its computational efficiency. The computational aspects of the extended FAST are described. These include (1) the definition of new sets of parametric equations for the search-curve exploring the input space, (2) the selection of frequencies for the parametric equations, and (3) the procedure adopted to estimate the total contributions. We also address the limitations of other global SA methods and suggest that the total-effect indices are ideally suited to perform a global, quantitative, model-independent sensitivity analysis.
Dynamic Bayesian networks (DBNs) are considered as a promising model for inferring gene networks from time series microarray
data. DBNs have overtaken Bayesian networks (BNs) as DBNs can construct cyclic regulations using time delay information. In
this paper, a general framework for DBN modelling is outlined. Both discrete and continuous DBN models are constructed systematically
and criteria for learning network structures are introduced from a Bayesian statistical viewpoint. This paper reviews the
applications of DBNs over the past years. Real data applications for Saccharomyces cerevisiae time series gene expression data are also shown.
Models and Sensitivity AnalysisMethods and Settings for Sensitivity Analysis – an IntroductionNonindependent Input FactorsPossible Pitfalls for a Sensitivity AnalysisConcluding RemarksExercisesAnswersAdditional ExercisesSolutions to Additional Exercises
Robustness is a ubiquitously observed property of biological systems. It is considered to be a fundamental feature of complex evolvable systems. It is attained by several underlying principles that are universal to both biological organisms and sophisticated engineering systems. Robustness facilitates evolvability and robust traits are often selected by evolution. Such a mutually beneficial process is made possible by specific architectural features observed in robust systems. But there are trade-offs between robustness, fragility, performance and resource demands, which explain system behaviour, including the patterns of failure. Insights into inherent properties of robust systems will provide us with a better understanding of complex diseases and a guiding principle for therapy design.