Advisor Information
Donald Rowen
Presentation Type
Poster
Start Date
1-3-2019 9:00 AM
End Date
1-3-2019 10:15 AM
Abstract
Properties emerge from the dynamics of large-scale molecular networks that are not discernible at the individual gene or protein level. Mathematical models - such as probabilistic Boolean networks - of molecular systems offer a deeper insight into how these emergent properties arise. Here, we introduce a non-linear, deterministic Boolean model of protein, gene, and chemical interactions in human macrophage cells during HIV infection. Our model is composed of 713 nodes with 1583 interactions between nodes and is responsive to 38 different inputs including signaling molecules, bacteria, viruses, and HIV viral particles. Additionally, the model accurately simulates the dynamics of over 50 different known phenomena, including molecular events associated with viral infection, endocytosis, transport, replication, budding, and cellular release. Statistical analyses of the model reveal network components with significant potential to influence molecular systems in both normal and infected macrophages, many of which have been confirmed in cell and animal models of HIV infection. We designed a Probabilistic Confidence Interval analysis for Boolean models (PCIB), demonstrating that our model emulates approximately 82% of a mass spectrometry dataset, collected from 7 macrophage samples infected with HIV across 67 proteins known to be central to the HIV infection process. The reproducibility of our model will facilitate guided hypothesis creation for future in vitro and in vivo studies. Additionally, the model allows for protein signaling interactions in human macrophages during HIV infection to be studied from a non-reductionist point of view.
Included in
Biochemistry Commons, Immunology and Infectious Disease Commons, Molecular Biology Commons
Large Scale Dynamical Model of Macrophage/HIV Interactions
Properties emerge from the dynamics of large-scale molecular networks that are not discernible at the individual gene or protein level. Mathematical models - such as probabilistic Boolean networks - of molecular systems offer a deeper insight into how these emergent properties arise. Here, we introduce a non-linear, deterministic Boolean model of protein, gene, and chemical interactions in human macrophage cells during HIV infection. Our model is composed of 713 nodes with 1583 interactions between nodes and is responsive to 38 different inputs including signaling molecules, bacteria, viruses, and HIV viral particles. Additionally, the model accurately simulates the dynamics of over 50 different known phenomena, including molecular events associated with viral infection, endocytosis, transport, replication, budding, and cellular release. Statistical analyses of the model reveal network components with significant potential to influence molecular systems in both normal and infected macrophages, many of which have been confirmed in cell and animal models of HIV infection. We designed a Probabilistic Confidence Interval analysis for Boolean models (PCIB), demonstrating that our model emulates approximately 82% of a mass spectrometry dataset, collected from 7 macrophage samples infected with HIV across 67 proteins known to be central to the HIV infection process. The reproducibility of our model will facilitate guided hypothesis creation for future in vitro and in vivo studies. Additionally, the model allows for protein signaling interactions in human macrophages during HIV infection to be studied from a non-reductionist point of view.