Predicting and programming microbial communities across scales
Duke University
Abstract:
A major goal in microbiome engineering is to identify molecular and ecological control knobs that can steer microbial communities toward desired functional states. Because microbial interactions are complex, nonlinear, and difficult to describe from first principles, computational modeling has become an indispensable tool in this field. To accelerate design–test–learn cycles in microbiome engineering and discover the mechanisms driving community dynamics, I will present new computational frameworks for physics-integrated machine learning and genome-to-function mapping.
Our physics-integrated machine learning models combine the flexibility of data-driven approaches with the interpretability of mechanistic models, achieving superior predictive performance compared to either alone. Moreover, these models can uncover causal relationships between species and metabolites, providing mechanistic insight into community dynamics. We demonstrate the ability to use active learning to efficiently navigate large experimental design landscapes with the model to fill knowledge gaps and optimize desired objectives. Complementarily, our genome-to-function modeling framework predicts the functional impact of previously unseen species, expanding the predictive power of synthetic community design.
Together, these tools enable the discovery of key molecular and ecological levers that control community behavior and will accelerate our ability to engineer microbial consortia with tailored functions for applications in precision medicine, sustainable agriculture, bioprocessing, environmental remediation, and the built environment.
Biography:
Dr. Ophelia Venturelli is an Associate Professor of Biomedical Engineering at Duke University. The Venturelli lab aims to understand and engineer microbiomes using systems and synthetic biology for applications spanning human health, agriculture and bioprocessing. Dr. Venturelli began as an Assistant Professor at UW-Madison in Biochemistry after completing a Life Sciences Research Foundation Fellowship at UC Berkeley. Dr. Venturelli’s postdoctoral research focused on developing data-driven methods to decipher microbial interactions shaping assembly of synthetic human gut microbiomes and strategies to manipulate intracellular resource allocation by exploiting tools from synthetic biology. She received her PhD in Biochemistry and Biophysics in 2013 from Caltech, where she studied single-cell dynamics and the role of feedback loops in a metabolic gene regulatory network. Dr. Venturelli received numerous awards including Shaw Scientist Award (2017), Army Research Office Young Investigator Award (2017), the Wisconsin Alumni Research Foundation Innovation Award (2019), OVCRGE Early Career Innovator Award (2023), ACS Synthetic Biology Young Investigator Award (2023) and the Thomas Langford Lectureship Award (2024).