+DS IPLE: Machine Learning for Synthetic and Quantitative Biology

A central objective in synthetic biology is to control the dynamics of engineered cells or cell populations in a predictable manner. Achieving this objective requires a quantitative description of biological systems that are both reliable and can be solved fast enough to guide experiments. To date, this line of work has primarily relied on the use of kinetic models that account for the relevant interactions. For instance, differential equations can be formulated to describe how bacterial populations form self-organized patterns or respond to different environmental cues, such as antibiotic treatment. However, these models face different conceptual and technical limitations, depending on the specific application contexts. They may fail to capture the relevant complexity of the experimental system. Or they may be computationally prohibitive to solve when one attempts to explore a large parametric space. I will discuss specific examples where machine learning can be applied to overcome these limitations. The combination of mechanistic modeling and machine learning can lead to computational predictions that are both effective and interpretable.

This session is part of the Duke+Data Science (+DS) program in-person learning experiences (IPLEs). To learn more, please visit https://plus.datascience.duke.edu/