Control of Medical Digital Twins with Artificial Neural Networks

The objective of personalized medicine is to tailor interventions to an individual patient's unique characteristics. A key technology for this purpose involves medical digital twins, computational models of human biology that can be personalized and dynamically updated to incorporate patient-specific data collected over time. Certain aspects of human biology, such as the immune system, are not easily captured with physics-based models, such as differential equations. Instead, they are often multi-scale, stochastic, and hybrid. This poses a challenge to existing model-based control and optimization approaches that cannot be readily applied to such models. Recent advances in automatic differentiation and neural-network control methods hold promise in addressing complex control problems. However, the application of these approaches to biomedical systems is still in its early stages. This work introduces dynamics-informed neural-network controllers as an alternative approach to control of medical digital twins. As a first use case for this method, the focus is on agent-based models, a versatile and increasingly common modeling platform in biomedicine. The effectiveness of the proposed neural-network control method is illustrated and benchmarked against other methods with two widely-used agent-based model types. The relevance of the method introduced here extends beyond medical digital twins to other complex dynamical systems.

Control of Medical Digital Twins with Artificial Neural Networks

The objective of personalized medicine is to tailor interventions to an individual patient's unique characteristics. A key technology for this purpose involves medical digital twins, computational models of human biology that can be personalized and dynamically updated to incorporate patient-specific data collected over time. Certain aspects of human biology, such as the immune system, are not easily captured with physics-based models, such as differential equations. Instead, they are often multi-scale, stochastic, and hybrid. This poses a challenge to existing model-based control and optimization approaches that cannot be readily applied to such models. Recent advances in automatic differentiation and neural-network control methods hold promise in addressing complex control problems. However, the application of these approaches to biomedical systems is still in its early stages. This work introduces dynamics-informed neural-network controllers as an alternative approach to control of medical digital twins. As a first use case for this method, the focus is on agent-based models, a versatile and increasingly common modeling platform in biomedicine. The effectiveness of the proposed neural-network control method is illustrated and benchmarked against other methods with two widely-used agent-based model types. The relevance of the method introduced here extends beyond medical digital twins to other complex dynamical systems.

Prospective predictors of electronic nicotine delivery system initiation in tobacco naive young adults: A machine learning approach

The use of electronic nicotine delivery systems (ENDS) is increasing among young adults. However, there are few studies regarding predictors of ENDS initiation in tobacco-naive young adults. Identifying the risk and protective factors of ENDS initiation that are specific to tobacco-naive young adults will enable the creation of targeted policies and prevention programs. This study used machine learning (ML) to create predictive models, identify risk and protective factors for ENDS initiation for tobacco-naive young adults, and the relationship between these predictors and the prediction of ENDS initiation. We used nationally representative data of tobacco-naive young adults in the U.S drawn from the Population Assessment of Tobacco and Health (PATH) longitudinal cohort survey. Respondents were young adults (18-24 years) who had never used any tobacco products in Wave 4 and who completed Waves 4 and 5 interviews. ML techniques were used to create models and determine predictors at 1-year follow-up from Wave 4 data. Among the 2,746 tobacco-naive young adults at baseline, 309 initiated ENDS use at 1-year follow-up. The top five prospective predictors of ENDS initiation were susceptibility to ENDS, increased days of physical exercise specifically designed to strengthen muscles, frequency of social media use, marijuana use and susceptibility to cigarettes. This study identified previously unreported and emerging predictors of ENDS initiation that warrant further investigation and provided comprehensive information on the predictors of ENDS initiation. Furthermore, this study showed that ML is a promising technique that can aid ENDS monitoring and prevention programs.

Computational models in systems biology: standards, dissemination, and best practices

Mathematical and computational models are a key technology in systems biology. Progress in the field depends on the replicability and reproducibility of their properties and behavior. For this, an essential requirement is a set of clear standards for model specification and dissemination. This review covers existing standards, and it highlights the most important areas where further work is required. This includes the specification of agent-based models, an increasingly common modeling approach.

E-Cigarette Use in Young Adult Never Cigarette Smokers with Disabilities: Results from the Behavioral Risk Factor Surveillance System Survey

Young adult never cigarette smokers with disabilities may be at particular risk for adopting e-cigarettes, but little attention has been paid to these people. This study examines the associations between different types of disability and e-cigarette use in this population. Young adult never-smokers from the 2016-2017 Behavioral Risk Factor Surveillance System (BRFSS) survey who were either never or current e-cigarette users (n = 79,177) were selected for the analysis. The Least Absolute Shrinkage and Selection Operator (LASSO) algorithm was used to select confounders for multivariable logistic regression models. Multivariable logistic regression models were used to determine the associations between current e-cigarette use and different types of disability after incorporating BRFSS survey design and adjusting for confounders. Young adult never-smokers who reported any disability had increased odds (OR 1.44, 95% CI 1.18-1.76) of e-cigarette use compared to those who reported no disability. Young adult never-smokers who reported self-care, cognitive, vision, and independent living disabilities had higher odds of e-cigarette use compared to those who reported no disability. There was no statistically significant difference in the odds of e-cigarette use for those reporting hearing and mobility disabilities compared to those who reported no disability. This study highlights the need for increased public education and cessation programs for this population.

Factors Associated with E-Cigarette Use in U.S. Young Adult Never Smokers of Conventional Cigarettes: A Machine Learning Approach

E-cigarette use is increasing among young adult never smokers of conventional cigarettes, but the awareness of the factors associated with e-cigarette use in this population is limited. The goal of this work was to use machine learning (ML) algorithms to determine the factors associated with current e-cigarette use among US young adult never cigarette smokers. Young adult (18-34 years) never cigarette smokers from the 2016 and 2017 Behavioral Risk Factor Surveillance System (BRFSS) who reported current or never e-cigarette use were used for the analysis (n = 79,539). Variables associated with current e-cigarette use were selected by two ML algorithms (Boruta and Least absolute shrinkage and selection operator (LASSO)). Odds ratios were calculated to determine the association between e-cigarette use and the variables selected by the ML algorithms, after adjusting for age, gender and race/ethnicity and incorporating the BRFSS complex design. The prevalence of e-cigarette use varied across states. Factors previously reported in the literature, such as age, race/ethnicity, alcohol use, depression, as well as novel factors associated with e-cigarette use, such as disabilities, obesity, history of diabetes and history of arthritis were identified. These results can be used to generate further hypotheses for research, increase public awareness and help provide targeted e-cigarette education.

Systems biology of ferroptosis: A modeling approach

Ferroptosis is a recently discovered form of iron-dependent regulated cell death (RCD) that occurs via peroxidation of phospholipids containing polyunsaturated fatty acid (PUFA) moieties. Activating this form of cell death is an emerging strategy in cancer treatment. Because multiple pathways and molecular species contribute to the ferroptotic process, predicting which tumors will be sensitive to ferroptosis is a challenge. We thus develop a mathematical model of several critical pathways to ferroptosis in order to perform a systems-level analysis of the process. We show that sensitivity to ferroptosis depends on the activity of multiple upstream cascades, including PUFA incorporation into the phospholipid membrane, and the balance between levels of pro-oxidant factors (reactive oxygen species, lipoxogynases) and antioxidant factors (GPX4). We perform a systems-level analysis of ferroptosis sensitivity as an outcome of five input variables (ACSL4, SCD1, ferroportin, transferrin receptor, and p53) and organize the resulting simulations into 'high' and 'low' ferroptosis sensitivity groups. We make a novel prediction corresponding to the combinatorial requirements of ferroptosis sensitivity to SCD1 and ACSL4 activity. To validate our prediction, we model the ferroptotic response of an ovarian cancer stem cell line following single- and double-knockdown of SCD1 and ACSL4. We find that the experimental outcomes are consistent with our simulated predictions. This work suggests that a systems-level approach is beneficial for understanding the complex combined effects of ferroptotic input, and in predicting cancer susceptibility to ferroptosis.

Sensitivity of comorbidity network analysis

OBJECTIVES

Comorbidity network analysis (CNA) is a graph-theoretic approach to systems medicine based on associations revealed from disease co-occurrence data. Researchers have used CNA to explore epidemiological patterns, differentiate populations, characterize disorders, and more; but these techniques have not been comprehensively evaluated. Our objectives were to assess the stability of common CNA techniques.

MATERIALS AND METHODS

We obtained seven co-occurrence data sets, most from previous CNAs, coded using several ontologies. We constructed comorbidity networks under various modeling procedures and calculated summary statistics and centrality rankings. We used regression, ordination, and rank correlation to assess these properties' sensitivity to the source of data and construction parameters.

RESULTS

Most summary statistics were robust to variation in link determination but somewhere sensitive to the association measure. Some more effectively than others discriminated among networks constructed from different data sets. Centrality rankings, especially among hubs, were somewhat sensitive to link determination and highly sensitive to ontology. As multivariate models incorporated additional effects, comorbid associations among low-prevalence disorders weakened while those between high-prevalence disorders shifted negative.

DISCUSSION

Pairwise CNA techniques are generally robust, but some analyses are highly sensitive to certain parameters. Multivariate approaches expose additional conceptual and technical limitations to the usual pairwise approach.

CONCLUSION

We conclude with a set of recommendations we believe will help CNA researchers improve the robustness of results and the potential of follow-up research.

Control of Intracellular Molecular Networks Using Algebraic Methods

Many problems in biology and medicine have a control component. Often, the goal might be to modify intracellular networks, such as gene regulatory networks or signaling networks, in order for cells to achieve a certain phenotype, what happens in cancer. If the network is represented by a mathematical model for which mathematical control approaches are available, such as systems of ordinary differential equations, then this problem might be solved systematically. Such approaches are available for some other model types, such as Boolean networks, where structure-based approaches have been developed, as well as stable motif techniques. However, increasingly many published discrete models are mixed-state or multistate, that is, some or all variables have more than two states, and thus the development of control strategies for multistate networks is needed. This paper presents a control approach broadly applicable to general multistate models based on encoding them as polynomial dynamical systems over a finite algebraic state set, and using computational algebra for finding appropriate intervention strategies. To demonstrate the feasibility and applicability of this method, we apply it to a recently developed multistate intracellular model of E2F-mediated bladder cancerous growth and to a model linking intracellular iron metabolism and oncogenic pathways. The control strategies identified for these published models are novel in some cases and represent new hypotheses, or are supported by the literature in others as potential drug targets. Our Macaulay2 scripts to find control strategies are publicly available through GitHub at https://github.com/luissv7/multistatepdscontrol.

A mathematical model of combined CD8 T cell costimulation by 4-1BB (CD137) and OX40 (CD134) receptors

Combined agonist stimulation of the TNFR costimulatory receptors 4-1BB (CD137) and OX40(CD134) has been shown to generate supereffector CD8 T cells that clonally expand to greater levels, survive longer, and produce a greater quantity of cytokines compared to T cells stimulated with an agonist of either costimulatory receptor individually. In order to understand the mechanisms for this effect, we have created a mathematical model for the activation of the CD8 T cell intracellular signaling network by mono- or dual-costimulation. We show that supereffector status is generated via downstream interacting pathways that are activated upon engagement of both receptors, and in silico simulations of the model are supported by published experimental results. The model can thus be used to identify critical molecular targets of T cell dual-costimulation in the context of cancer immunotherapy.

A Systems Biology Approach to Understanding the Pathophysiology of High-Grade Serous Ovarian Cancer: Focus on Iron and Fatty Acid Metabolism

Ovarian cancer (OVC) is the most lethal of the gynecological malignancies, with diagnosis often occurring during advanced stages of the disease. Moreover, a majority of cases become refractory to chemotherapeutic approaches. Therefore, it is important to improve our understanding of the molecular dependencies underlying the disease to identify novel diagnostic and precision therapeutics for OVC. Cancer cells are known to sequester iron, which can potentiate cancer progression through mechanisms that have not yet been completely elucidated. We developed an algorithm to identify novel links between iron and pathways implicated in high-grade serous ovarian cancer (HGSOC), the most common and deadliest subtype of OVC, using microarray gene expression data from both clinical sources and an experimental model. Using our approach, we identified several links between fatty acid (FA) and iron metabolism, and subsequently developed a network for iron involvement in FA metabolism in HGSOC. FA import and synthesis pathways are upregulated in HGSOC and other cancers, but a link between these processes and iron-related genes has not yet been identified. We used the network to derive hypotheses of specific mechanisms by which iron and iron-related genes impact and interact with FA metabolic pathways to promote tumorigenesis. These results suggest a novel mechanism by which iron sequestration by cancer cells can potentiate cancer progression, and may provide novel targets for use in diagnosis and/or treatment of HGSOC.

Abstract 1322: A systems-level approach identifies novel links between iron and fatty acid metabolism in high-grade serous ovarian cancer

The low survival rate for high-grade serous ovarian cancer (HGSOC) motivates novel approaches to identify causative and therapeutically targetable factors involved in HGSOC initiation and progression. Cancer cells are known to sequester iron, which can potentiate cancer progression via mechanisms that have not yet been completely elucidated. In order to uncover novel, and potentially therapeutically tractable links, between iron and HGSOC, we developed an algorithm to identify perturbed regulatory pathways containing differentially expressed iron-related genes in microarray data from clinical sources and an experimental model of HGSOC. Used in tandem with over-representation analysis for iron-related genes, the algorithm led us to uncover an iron dependence of fatty acid import and synthesis pathways, which are upregulated in HGSOC and other cancers, and to develop a network synthesizing the relationship between iron, fatty acid metabolism, and HGSOC. We use the network to derive specific hypotheses of mechanisms by which iron impacts fatty acid metabolic pathways to promote tumorigenesis. We have thus shown that a systems-level approach to identifying novel regulatory links between iron and HGSOC has yielded a previously unappreciated association between iron and fatty acid metabolism that may be exploited for therapeutic potential. This work has been supported in part by NIH grants F32CA214030 to AK, R01CA188025 to SVT, and RO1CA171101 to FMT.

Citation Format: Anna Konstorum, Miranda L. Lynch, Suzy V. Torti, Frank M. Torti, Reinhard C. Laubenbacher. A systems-level approach identifies novel links between iron and fatty acid metabolism in high-grade serous ovarian cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1322.

Addressing current challenges in cancer immunotherapy with mathematical and computational modelling

The goal of cancer immunotherapy is to boost a patient's immune response to a tumour. Yet, the design of an effective immunotherapy is complicated by various factors, including a potentially immunosuppressive tumour microenvironment, immune-modulating effects of conventional treatments and therapy-related toxicities. These complexities can be incorporated into mathematical and computational models of cancer immunotherapy that can then be used to aid in rational therapy design. In this review, we survey modelling approaches under the umbrella of the major challenges facing immunotherapy development, which encompass tumour classification, optimal treatment scheduling and combination therapy design. Although overlapping, each challenge has presented unique opportunities for modellers to make contributions using analytical and numerical analysis of model outcomes, as well as optimization algorithms. We discuss several examples of models that have grown in complexity as more biological information has become available, showcasing how model development is a dynamic process interlinked with the rapid advances in tumour-immune biology. We conclude the review with recommendations for modellers both with respect to methodology and biological direction that might help keep modellers at the forefront of cancer immunotherapy development.

Applications of network analysis to routinely collected health care data: a systematic review

Objective

To survey network analyses of datasets collected in the course of routine operations in health care settings and identify driving questions, methods, needs, and potential for future research.

Materials and Methods

A search strategy was designed to find studies that applied network analysis to routinely collected health care datasets and was adapted to 3 bibliographic databases. The results were grouped according to a thematic analysis of their settings, objectives, data, and methods. Each group received a methodological synthesis.

Results

The search found 189 distinct studies reported before August 2016. We manually partitioned the sample into 4 groups, which investigated institutional exchange, physician collaboration, clinical co-occurrence, and workplace interaction networks. Several robust and ongoing research programs were discerned within (and sometimes across) the groups. Little interaction was observed between these programs, despite conceptual and methodological similarities.

Discussion

We use the literature sample to inform a discussion of good practice at this methodological interface, including the concordance of motivations, study design, data, and tools and the validation and standardization of techniques. We then highlight instances of positive feedback between methodological development and knowledge domains and assess the overall cohesion of the sample.

Effects of research complexity and competition on the incidence and growth of coauthorship in biomedicine

BACKGROUND

Investigations into the factors behind coauthorship growth in biomedical research have mostly focused on specific disciplines or journals, and have rarely controlled for factors in combination or considered changes in their effects over time. Observers often attribute the growth to the increasing complexity or competition (or both) of research practices, but few attempts have been made to parse the contributions of these two likely causes.

OBJECTIVES

We aimed to assess the effects of complexity and competition on the incidence and growth of coauthorship, using a sample of the biomedical literature spanning multiple journals and disciplines.

METHODS

Article-level bibliographic data from PubMed were combined with publicly available bibliometric data from Web of Science and SCImago over the years 1999-2007. We selected four predictors of coauthorship were selected, two (study type, topical scope of the study) associated with complexity and two (financial support for the project, popularity of the publishing journal) associated with competition. A negative binomial regression model was used to estimate the effects of each predictor on coauthorship incidence and growth. A second, mixed-effect model included the journal as a random effect.

RESULTS

Coauthorship increased at about one author per article per decade. Clinical trials, supported research, and research of broader scope produced articles with more authors, while review articles credited fewer; and more popular journals published higher-authorship articles. Incidence and growth rates varied widely across journals and were themselves uncorrelated. Most effects remained statistically discernible after controlling for the publishing journal. The effects of complexity-associated factors held constant or diminished over time, while competition-related effects strengthened. These trends were similar in size but not discernible from subject-specific subdata.

CONCLUSIONS

Coauthorship incidence rates are multifactorial and vary with factors associated with both complexity and competition. Coauthorship growth is likewise multifactorial and increasingly associated with research competition.