Robust Approaches to Generating Reliable Predictive Models in Systems Biology
A computational technique is described to reduce the model search space and construct an ensemble of models for systems biology using perturbation data. While doing so, an effective way of representing a network model is developed for computing purposes using adjacency matrix-like data structures. This allows models to include Uni-Uni to Bi-Bi reactions in addition to enzymatic activation and inhibition. It is demonstrated that the technique is effective, fast, and suggests it can be used as an initial filtering step in conjunction with other computational techniques. Finally, other potential methods to construct a set of reliable network models using time-course data are explored.
KeywordsSystems biology Biochemical networks Network reduction Machine learning Ensemble modeling
KC is supported by NIH grants GM123032-01A1. The content is solely the responsibility of the author and does not necessarily represent the views of the National Institutes of Health. KC wishes to thank Herbert Sauro and Joseph Hellerstein for their help and guidance in completing this chapter.
- Abadi M, Agarwal A, Barham P et al (2015) Tensorflow: large-scale machine learning on heterogeneous distributed systems. http://download.tensorflow.org/paper/whitepaper2015.pdf
- Choi K, Medley JK, Cannistra C et al (2016) Tellurium: a python based modeling and reproducibility platform for systems biology. bioRxiv p 054601. https://doi.org/10.1101/054601
- Chollet F et al (2015) Keras. https://github.com/fchollet/keras
- Natale JL, Hofmann D, Hernández DG et al (2017) Reverse-engineering biological networks from large data sets. arXiv preprint arXiv:170506370Google Scholar
- Pedregosa F, Varoquaux G, Gramfort A et al (2011) Scikit-learn: machine learning in python. J Mach Learn Res 12:2825–2830Google Scholar
- Shi T, Niepel M, McDermott JE et al (2016) Conservation of protein abundance patterns reveals the regulatory architecture of the EGFR-MAPK pathway. Sci Signal 9(436):rs6Google Scholar