Introduction: Cancer Gene Networks
Robert Clarke
DOI: https://doi.org/10.1007/978-1-4939-6539-7_1
Abstract:Constructing, evaluating, and interpreting gene networks generally sits within the broader field of systems biology, which continues to emerge rapidly, particular with respect to its application to understanding the complexity of signaling in the context of cancer biology. For the purposes of this volume, we take a broad definition of systems biology. Considering an organism or disease within an organism as a system, systems biology is the study of the integrated and coordinated interactions of the network(s) of genes, their variants both natural and mutated (e.g., polymorphisms, rearrangements, alternate splicing, mutations), their proteins and isoforms, and the organic and inorganic molecules with which they interact, to execute the biochemical reactions (e.g., as enzymes, substrates, products) that reflect the function of that system. Central to systems biology, and perhaps the only approach that can effectively manage the complexity of such systems, is the building of quantitative multiscale predictive models. The predictions of the models can vary substantially depending on the nature of the model and its inputoutput relationships. For example, a model may predict the outcome of a specific molecular reaction(s), a cellular phenotype (e.g., alive, dead, growth arrest, proliferation, and motility), a change in the respective prevalence of cell or subpopulations, a patient or patient subgroup outcome(s). Such models necessarily require computers. Computational modeling can be thought of as using machine learning and related tools to integrate the very high dimensional data generated from modern, high throughput omics technologies including genomics (next generation sequencing), transcriptomics (gene expression microarrays; RNAseq), metabolomics and proteomics (ultra high performance liquid chromatography, mass spectrometry), and "subomic" technologies to study the kinome, methylome, and others. Mathematical modeling can be thought of as the use of ordinary differential equations and related tools to create dynamic, semi-mechanistic models of low dimensional data including gene/protein signaling as a function of time/dose. More recently, the integration of imaging technologies into predictive multiscale modeling has begun to extend further the scales across which data can be obtained and used to gain insight into system function.There are several goals for predictive multiscale modeling including the more academic pursuit of understanding how the system or local feature thereof is regulated or functions, to the more practical or translational goals of identifying predictive (selecting which patient should receive which drug/therapy) or prognostic (disease progress and outcome in an individual patient) biomarkers and/or identifying network vulnerabilities that represent potential targets for therapeutic benefit with existing drugs (including drug repurposing) or for the development of new drugs. These various goals are not necessarily mutually exclusive or inclusive. Within this volume, readers will find examples of many of the activities noted above. Each chapter contains practical and/or methodological insights to guide readers in the design and interpretation of their own and published work.