Modeling spatial interaction networks of the gut microbiota
Xiaocang Cao,
Ang Dong,
Guangbo Kang,
Xiaoli Wang,
Liyun Duan,
Huixing Hou,
Tianming Zhao,
Shuang Wu,
Xinjuan Liu,
He Huang,
Rongling Wu
Affiliations
Xiaocang Cao
Department of Gastroenterology and Hepatology, Tianjin Medical University General Hospital, Tianjin Medical University, Tianjin, China
Ang Dong
Center for Computational Biology, College of Biological Sciences and Technology, Beijing Forestry University, Beijing, China
Guangbo Kang
School of Chemical Engineering and Technology, Frontiers Science Center for Synthetic Biology and Key Laboratory of Systems Bioengineering (Ministry of Education), Tianjin University, Tianjin, China
Xiaoli Wang
Department of Gastroenterology and Hepatology, Tianjin Medical University General Hospital, Tianjin Medical University, Tianjin, China
Liyun Duan
Department of Gastroenterology and Hepatology, Tianjin Medical University General Hospital, Tianjin Medical University, Tianjin, China
Huixing Hou
Department of Gastroenterology and Hepatology, Tianjin Medical University General Hospital, Tianjin Medical University, Tianjin, China
Tianming Zhao
Department of Gastroenterology and Hepatology, Tianjin Medical University General Hospital, Tianjin Medical University, Tianjin, China
Shuang Wu
Center for Computational Biology, College of Biological Sciences and Technology, Beijing Forestry University, Beijing, China
Xinjuan Liu
Department of Gastroenterology, Beijing Chaoyang Hospital, Capital Medical University, Beijing, China
He Huang
School of Chemical Engineering and Technology, Frontiers Science Center for Synthetic Biology and Key Laboratory of Systems Bioengineering (Ministry of Education), Tianjin University, Tianjin, China
Rongling Wu
Center for Statistical Genetics, Departments of Public Health Sciences and Statistics, The Pennsylvania State University, Hershey, PA, USA
How the gut microbiota is organized across space is postulated to influence microbial succession and its mutualistic relationships with the host. The lack of dynamic or perturbed abundance data poses considerable challenges for characterizing the spatial pattern of microbial interactions. We integrate allometric scaling theory, evolutionary game theory, and prey-predator theory into a unified framework under which quasi-dynamic microbial networks can be inferred from static abundance data. We illustrate that such networks can capture the full properties of microbial interactions, including causality, the sign of the causality, strength, and feedback loop, and are dynamically adaptive along spatial gradients, and context-specific, characterizing variability between individuals and within the same individual across time and space. We design and conduct a gut microbiota study to validate the model, characterizing key spatial determinants of the microbial differences between ulcerative colitis and healthy controls. Our model provides a sophisticated means of unraveling a complete atlas of how microbial interactions vary across space and quantifying causal relationships between such spatial variability and change in health state.
