Disentangling intrinsic and extrinsic stochasticity to understand epigenetic patterning and stability

Mathematical modeling plays an increasingly important role in understanding cell biological processes and in bioengineering. Models that can account for “stochasticity”, or fluctuations that occur at the molecular level, can help make sense of noisy data and predict system behavior. Improved mathematical models of stochasticity in cell biology could impact broad areas, from medicine to agriculture to industrial biotechnology. This project will develop new methods for mathematical modeling and data analysis, with the goal of identifying and disentangling the different sources of noise that are present in modern genetic and epigenetic sequencing data. The project will focus initially on DNA methylation, a common epigenetic mechanism in plants, animals, and fungi. However, the newly developed methods could be useful in broad areas of science and engineering that are challenged by noisy data and low-time-resolution measurements. The project will also develop publicly available software tools and contribute to recruitment and training young scientists. A key project focus is on the problem of distinguishing between intrinsic and extrinsic stochasticity in biological data. This is an important challenge, because incorrect noise models can thwart interpretation of data. Also, understanding the interplay of various noise sources reveals fundamental limits on the functioning of biochemical systems. Specific project aims include the determination of how extrinsic variability hinders inference of dynamics from data, with a focus on multimodal stochastic systems relevant to epigenetic regulation and gene expression. Then, methods will be developed to separate extrinsic noise contributions, focusing on varying cellular and sub-cellular genomic environments. The newly developed methods will be used to predict intrinsically dynamic regions of the epigenome from sequencing data. Finally, model predictions will be validated using experimental data. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.