Our lab studies the statistical foundations of microbiome analysis. We develop tools to support precision and discovery from high-throughput microbial ecosystem data, and we draw inspiration from daily collaboration with researchers solving complex real-world problems.

We explore how data visualization, simulation, and representation learning can guide reasoning about data gathered from heterogeneous platforms. In the process, we critically examine widespread data analysis practices, engage with developments in statistics and machine learning, and build modular, user-centric software. Ultimately, we aim to facilitate fluid, formal, and imaginative data analysis in problems critical to human and planetary health.

• Semisynthetic Simulation for Microbiome Data Analysis
• Multiscale Analysis of Count Data through Topic Alignment
• multimedia: Multimodal Mediation Analysis of Microbiome Data
• Data Science Principles for Interpretable and Explainable AI
• Generative Models: An Interdisciplinary Perspective
• Latent Variable Modeling for the Microbiome

Code adapted from Slime Mold Simulation through a CC BY-NA-SA 3.0 license.