We propose to transform our CompuCell3D virtual tissue simulator into a bona-fide computer aided engineering (CAE) system for bio-medical research focused on the biologically-intuitive description and design of multi- cell models of tissues and organs (Virtual Tissues). This will include descriptions of biological mechanisms, components, processes and outcomes. Mechanism-based models of this type will allow researchers to explore the systemic outcomes of molecular and cellular perturbations (characterized using in vitro and high-throughput experiments) to extract mechanistic understanding of normal and disease states. CAE programs allow engineers to construct virtual representations (computational models) of components, assemble them into complex devices, and simulate their behaviors and interactions in virtual experiments. CAE programs allow engineers to evaluate the performance, reliability and failure modes of proposed devices without actually having to physically build those devices. CAE platforms enable the capture of domain knowledge in validated, sharable and reusable components and composite models and to leverage big-data to abstract knowledge, increasing the speed, efficiency and reliability of experiment and design. Significance: A key impediment to adoption of CAE simulation approaches in biology and medicine is not the speed or efficiency of simulation software, but rather the lack of appropriate human-computer interaction between life-science researchers and simulation software. Increasing the uptake of CAE tools in biomedicine requires the development of biologically motivated languages and interfaces that facilitate the creation of mechanism-based models of biological systems, and embed biological knowledge within these models in a way that promotes knowledge validation, mining, recovery and reuse. Innovation: We will create a platform for building modular, sharable and reusable virtual-tissue models (including patient-specific models that integrate patient-derived measurements) that capture and integrate biological and mechanistic knowledge and experimental observations. Building virtual-tissue models requires combining and integrating a range of components from different scales. We develop a new hybrid programming language based on biological rather than computational concepts that naturally expresses the complex multi-scale objects and dynamic interactions in a unified way. To our knowledge, this language is the first if its kind to allow biological components and models to be visually created and composite and enables the capture, search, formalization, extraction and reuse of domain knowledge. Our platform will transform these models into executable simulations, creating virtual experiments which allow direct comparison with laboratory experiments. We will also develop tools to optimize virtual-tissue models against qualitative and semi-quantitative experimental data, providing robust techniques to explore diverse hypotheses and parameter ranges, and optimally employ available experimental data. These approaches will reduce the effort required to develop understanding of normal and pathological conditions and provide a route to extraction of knowledge from heterogeneous biological big data.
Modern biomedical research generates a wealth of molecular data (genomics, proteomics, and metabolomics) in both animals and humans. Modern imaging techniques produce complementary large data sets. However, we lack methods that can integrate molecular and imaging data to provide mechanism-based predictions of human health outcomes. We will develop computational tools that facilitate the creation of computational models that couple data-based descriptions of molecular, cellular and tissue behavior with imaging information to create predictive quantitative simulations of how molecular and cell-scale perturbations lead to normal and disease phenotypes at tissue and organism scales.
| Somogyi, Endre (2018) A Dynamic Non-Manifold Mesh Data Structure to Represent Biological Materials. J WSCG (Plzen) 26:21-30 |
| Fu, Xiao; Sluka, James P; Clendenon, Sherry G et al. (2018) Modeling of xenobiotic transport and metabolism in virtual hepatic lobule models. PLoS One 13:e0198060 |
| Hagar, Amit; Flynn, Sean; Patterson, Katherine et al. (2018) Muscular endurance and progression rates of early-stage invasive ductal carcinoma: A pilot study. Breast J 24:849-851 |
| Somogyi, Endre; Glazier, James A (2017) A MODELING AND SIMULATION LANGUAGE FOR BIOLOGICAL CELLS WITH COUPLED MECHANICAL AND CHEMICAL PROCESSES. Symp Theory Model Simul 2017: |
| Somogyi, Endre; Sluka, James P; Glazier, James A (2016) Formalizing Knowledge in Multi-Scale Agent-Based Simulations. Model Driven Eng Lang Syst 16:115-122 |
| Somogyi, Endre; Hagar, Amit; Glazier, James A (2016) TOWARDS A MULTI-SCALE AGENT-BASED PROGRAMMING LANGUAGE METHODOLOGY. Proc Winter Simul Conf 2016:1230-1240 |
